ESSAY ON THE ROLE OF SENSATIONS IN ORDINARY LANGUAGE

W. has usefully emphasised a number of correct things about how ordinary language works:

1.  I did not learn our ordinary language in the case of, say, 'red' by my parents saying 'red while pointing at the sensation in the privacy of my mind. They pointed at an object, or a coloured shape in a book.
2. The public language does a useful job regardless of what sensation people have when they look at that sofa. "Bring us back some red apples" works, even if A has consistently got colour sensation inversion.
3. Even if some people are robots, or we live in a (near) future with helpful robots, we can still  say "Make sure to use red peppers" to the zombie or chef robot, and it will, if correctly designed, cook with the peppers we wanted it to, regardless of whether it is conscious or not - whether it has any sensations at all. This is the relevance of W.'s remark that any kind of beetle, or any other object, or even nothing, could be in our box. We could call the robot case a 'pure object language'.
4. A pure private language of pure sensations, not associated even conjecturally with public events, even if it could be instituted by a single person (which is doubtful), would be useless for communication between people. We could call this a 'pure sensation language'.
6. BUT:
8. These pure cases are model languages. They are very useful to have in mind, in order to clarify - to analyse - the complexity of our actual language, and they say almost everything that needs to be said, but not quite everything. For a complete description of ordinary language we need to combine these pure cases in just one more little step.
9. The key assumption that people make is: A: "Everyone's sensations are (broadly) similar when they are in the presence of similar objects". The characteristic of objects that we refer to with 'red' is, we assume, constantly correlated with a sensation - or, more precisely, the sensation that we have.
10. This assumption A appears to be entirely untestable.
11. People thus actually intend their ordinary language, which has the word 'sensation' in it, to have this dual (composite) nature. I can suppose that when I say "That is a nice red sunset" to Foad, I am communicating my pleasure at the private sensation, and expecting that if he looks at the sunset, he will have the same sensation. I am only actually doing this if A is true, and I cannot show that it is.
12. The door seems to open for some philosophising - but it doesn't open much. We cannot use language to describe our sensations; we can coherently ask the question "Does Matt have the same colour sensation as Edith when they look at that sofa?", but we can't answer it. That is just about it. But it is