Sunday, June 14, 2009

general notes on general semantics

One of the premises of general semantics is that our thoughts are controlled by the meanings that we assign to words. I'll quote from John C. Wright's introduction to "The Null-A Continuum":
The theory of general semantics postulates that through a proper understanding of the relationship between words and the reality words allegedly represent a mind can be trained to avoid disorientation. On an emotional level, a lack of disorientation means the absence of neurotic and self-destructive behavior.
This is reminiscent of some philosophers of the early twentieth century who tried to reduce thought to symbol manipulation. Symbol manipulation is just computation, and I explained in the previous post why computation cannot expand reasoning power. In this post, I will concentrate on three points: it is implausible that our thoughts really are limited by the words (or other symbols) that we have available, symbols cannot represent without the presence of a mental act which comes before the symbol, and if our thoughts really are limited by symbols then we have no way to judge which symbols are good ones.

I sometimes think in sentences. This happens when I am formulating an argument or trying out in my head something that I may write or say. When it does happen I am fully aware of it. Other times I think in terms of diagrams or physical geometry. I do this when I am trying to solve a problem that is too complex for my bare intuition to handle; I (like most people) have a more effective grasp of spatial relationships than of abstract relationships. Again, when I am thinking this way, I am fully aware of it. Most of the time, whether I am working hard at solving a design problem or wandering idly down the beach enjoying the sound of the waves, I don't think in symbols.

Some philosophers and psychologists would say that I am thinking in symbols but am just not aware of it. This view was especially prevalent during the first half of the twentieth century. But how can I be thinking in symbols when I am not aware of the symbols? Symbols are objects of consciousness. A symbol is a symbol in virtue of what it means to someone. Awareness of meaning is a conscious mental state, and without awareness there is no meaning. Being aware of a symbol unconsciously is a contradiction in terms.

Is there a more plausible to understanding of this idea? Maybe the underlying idea is that I have mental states that are the cognitive equivalent of symbols in the sense that they are physical objects or states in my brain that represent or are about something else. But how could physical states in my brain represent something else? It can't be the normal form of representation because as I argued before, something cannot represent in the usual way without a conscious intention for it to represent. The only physical way for something to represent would be with physical properties. What physical properties and relationships go into the relationship of representing?

One proposal is that representation is the physical relationship of similarity. For example a portrait represents a face by being visually similar to the face. A map represents the terrain by having elements that are geometrically similar to the terrain. But this account fails for two reasons. First, the notion of similarity is completely arbitrary. Does a glass of water represent Lake Michigan by way of chemical similarity? Does a piece of granite represent the Hoover Damn by way of having similar hardness? Even if you restrict "similarity" to being a geometric similarity it is peculiar that the lines on a map, which are just discolored sections of the page become representation of actual physical objects on the terrain. And what if John has a portrait taken but then grows a beard so that the portrait no longer looks like him at all? Does the portrait then come to represent his twin brother James who is beardless?

Clearly similarity does not work. Even if it worked in those cases that I mentioned (maps and portraits) it would not work for arbitrary symbols such as the word "Obama". The word "Obama" is not at all similar to the person the word represents. Representation is not a physical property at all, it is a mental property. A symbol represents in virtue of a mind intending for it to represent. And if a mind intends for the symbol S to represent an object O, then the mind must be able to form the concept of O without already having the relationship of symbolizing or it would never have formed the intention to represent O. If the mind needs a symbol to entertain a concept, then there is no place to get started. It needs a symbol to have a concept, but it needs a concept to have a symbol.

Minds can grasp concepts and think about concepts without symbols. And that is a good thing because if we needed symbols to think about things then we would not be able to grasp novel situations.

But suppose, against all evidence, our thoughts actually are controlled by words and what we think that they represent. How then could we possibly judge that some semantics is better than another? After all, if you are going to tell me that this relationship is better than that one then I need to use another such relationship between words and reality to decided if you are correct. But how do I know that that prior relationship is correct?

No comments: