↑ 21 References
This is also another form of the idea that definitions are clouds in concept-space, as a definition allows us to go up a layer of abstraction, and then create new definitions using other definitions. Layers of abstraction exist as layers of concept-space.
Another way of putting it: Proper naming conventions should give us a hint about how the properties of the named phenomena are distributed in concept-space.
However, even if there is ambiguity or uncertainty in a given section of concept-space, we should still seek to keep that words are pointers to concept-space stable. This allows us to avoid the fallacy of compression.
Quality understanding involves not only understanding how things are similar, but also how they are different. Understanding the details of things involves reveals their differences, and allows you to trim the edges of your concept-space.
When discussing a potentially difficult topic, it can be easy to have an idea be a poorly defined pointer to concept-space. This is inevitable, and why we use intentionally ambiguous naming to avoid restricting growing or uncertain ideas, for example.
Deductive logic manages definitions where inductive logic manages observations. We can use deductive logic to explore the rest of our concept-space for a given inductively attained belief. We can use the probability that our inductively attained belief is true to inform the deductively attained consequences of that belief. In turn, we can use those deductively attained consequences to provide an extension with which to further test our inductively attained belief. In a way, I guess this is science. science is the systematic use of full logic.
- concept-space: Our mental model that contains everything we think about. This would fall under the traditional idea of a priori thought. These are the mental frameworks we use to communicate with ourselves and others.
This is a disagreement between Concept-spaces, about definitions of words (since definitions are clouds in concept-space and words are pointers to concept-space). This type of argument likely does not have any possible external validation, and runs the risk of going on forever.
(One can get into Deductive thinking here, since words are pointers to concept-space and if one believes that the concept-space for “thoughts” and the set of all things I think are the exact same, or something like that. If we realize we are only thinking deductively, then it likely will serve us to focus on outcomes not definitions.)
Words are pointers to nebulously defined areas in concept-space. (Probably need to expand on this)
Since words are pointers to concept-space, we can see that definitions are areas of that concept-space.
This is a disagreement between Concept-spaces, about definitions of words (since definitions are clouds in concept-space and words are pointers to concept-space). This type of argument likely does not have any possible external validation, and runs the risk of going on forever.
An efficient ( optimization?) form of language should involve short messages that, through the use of nouns and adjectives, both communicate the ideas and imply their probabilistic relationships.