Intelligent Systems And Their Societies Walter Fritz

Growth Of A Concept


Growth Through Experience
Every time an IS has an experience that shows it something new about its environment, its brain creates a new concept or expands the corresponding, already existing concept. For instance, let's take the fresh mind of a young child and let that child experience seeing a cat and being told the word "cat" for the first time. From this, the child's brain forms a concept: "cat". Later on, the same child may have another experience where she/he sees a dog and, mistakenly -- to us that is -- says "cat". After all, this animal is also a furry being -- "fur is soft!" -- that "crawls fast" with four legs. However now it is told "dog". By noting the differences, the child's brain then amplifies its concept of cat by adding the short and wide head, the very long tail, the "meow" and "purr" sounds, etc. Later on, after experiences that allow him/her to get to know different types of cats, the brain will further amplify its concept by differentiating between different races of cats and probably even individual cats.


Convergence Of Growth
Eventually, the child will grow a concept of "cat" that is very similar, though not identical, to the cat concept its parents have. But why isn't it identical? After all, isn't "a cat is a cat", no matter who experiences it? The answer: Yes, and No.

Luckily, a simple explanation exists. It turns out that the difference between peoples' concepts stem from the fact that each person's experiences, and the order in which she/he has had these experiences are different. This difference in the experience process results in a slightly different, or "individualized" concept.


When has a concept grown sufficiently?
When a concept contains more information, we say that it is a "better" concept. For instance, a Greek philosopher once proposed that the "thing itself" is not an apple, but rather a collection of atoms. Today, a physicist would say that it is a collection of protons, neutrons, and electrons. Similarly, a nuclear physicist might say that it is a collection of quarks. What will scientists 2000 years from now say? Who knows?

Really, it doesn't matter. Regardless of who and when or even what they say (if they even "talk" at all), ISs of all kinds will still think in concepts. Even though 2000 years from now their concepts will undoubtedly be amplified well beyond those of today, we can still say that these concepts will only correspond to the "thing itself"; they are not the "thing itself" and indeed cannot completely identify the "thing itself". We will know more about the thing itself, but never all its properties.


Growth And Usefulness
Newton said that light was a particle; Huyghens said it was a wave. But "particle" and "wave" are concepts, not the "thing itself". Thus light, the "thing itself", is neither. Light is not the concept "particle" nor the concept "wave".

It turns out that we can fruitfully apply the concept wave in thinking about light and predicting its behavior in certain circumstances. Equally, we can use the concept "particle" (with all its connecting concepts) to predict the properties and actions of light in certain other circumstances. But light is still neither particle nor wave; light is an unknowable, a "thing itself".
When we finally put this all together, we find that:

  We can use concepts to "know" about how "things" act in our environment,  
but cannot use concepts to actually "know" the "thing itself".


For continuous reading, like a book - continue here.
Jump to the top of this document /Details of IS /Intelligent Systems /E-book Contents.

Last Edited 22 Julio 2013 / Walter Fritz
Copyright © New Horizons Press