Communication is that process by which events in the subjective world of an individual are represented in the objective world.
The function which maps from the perceptual domain into the objective domain is known as communication. Both symbolic and sub-symbolic information can be communicated to the external world, but this section focuses on the external representation of symbolic information. A language is necessary for symbolic communication, and to the extent that we identify thought with language, language is also necessary for thought.
Communication is complicated in part because the concepts of the speaker and the listener are not identical, even for identical words. It is also complicated if the listener does not know all of the words that the speaker uses. However, the former issue is the more insidious of the two problems, since the listener recognizes the words that the speaker is using, but associates them with different definitions. In light of this, the listener incorrectly believes that he or she understands the speaker's intended meaning.
Speakers of the same language have concepts that can be verified to apply to the same things (at least approximately). It is not clear, however, that their percepts are the same. For example, when you see an apple, and I see the same apple (from the same perspective), do we see the same thing? In other words, for identical objects, are our perceptions identical? There is no way to be sure. We might communicate with one another in order to arrive at an answer. Using communication to verify this hypothesis, however, causes at least two problems: the first difficulty is the mapping between the apple object and the ‘apple’ percept (which occurs internally for each of us), and the second difficulty is the mapping between the ‘apple’ percept and the “apple” concept (without which we would not be able to communicate our experience of the apple). Both of these problems are captured in the following more generic question: is it possible to ensure that we are having the same subjective experience upon seeing the same apple?
Between referential domains, the only available conditions for identity are those of isomorphism.
If you and I have percepts of an apple that are isomorphic to one another, as well as concepts and symbols that are isomorphic to these percepts, then we will not be able to tell from our descriptions if we are having the same experience when viewing the same thing. We will both say that we see an apple. If the apple is red, then we will both verbally agree on that point. However, this agreement about the redness of the apple does not help to answer the question; it only transforms the question to “When we see a red thing, does the redness appear the same to both of us?” . This train of thought merely transforms the original question into other related questions, in a circular way. Ultimately, all we can be sure of is that the relations between the words of the language that we speak are the same (i.e. that our lexicons are isomorphic). Identity, as opposed to isomorphism, cannot be achieved in this way.
To carry this thought experiment a bit further, suppose that our perceptions of green and red are switched. In other words, what you see when you look at a red thing looks like what I see when I look at a green thing. This does not mean that you are allowed to run lights while driving: “Red means stop” still applies to both of us. Similarly, we both call the same things red: when we learn language, we learn to link a word with our subjective experience of that thing, whatever that subjective experience may be .
Now suppose that I could somehow directly perceive your percepts when you look at an apple: would I recognize that object as an apple? Is it possible that it looks, to me, like what I see when I look at an orange? This seems unlikely at first, but we cannot rely on color to distinguish between the two since orange and red might be switched. Similarly, various other properties (in addition to color) might be flip-flopped for us. So how can we be sure? This problem is exacerbated by the fact that we do not need to have a real object present in order to generate the percept in us (e.g. as happens when we dream or hallucinate).[87]
This situation is graphically depicted in the next two figures. In the following figure, we have depicted the objects, the ‘percepts’ , and the “concepts” that correspond to an apple and an orange for a particular individual.
In the following figure, the perceptions of apple and orange are switched. The difficulty in distinguishing between these two diagrams is that, at the conceptual level, there is no observable difference in the mapping from objects to concepts. In both cases, the concepts correctly map to their intended objects. The difficulty is that because we cannot directly perceive the percepts of the individual, we have no way to know the organization of the internal mapping. In other words, the middle layer of perception cannot be seen by an external observer: it is a hidden layer.
[87] For vision in particular, this is a poor example because we know that spatial mapping is preserved well into the occipital lobe. In general, apples and oranges have a number of topological features which probably tightly constrain their representation in the brain.