Another way of looking at the problem is to ask: If I wake up one morning, and my spectrum is inverted, what happens? Will I get used to it over time? Let's say that I see "red" things, like apples and blood, as green. (The meaning of this as is something to ponder.) After a few weeks, I am able to correctly apply the term "red" to apples and blood. Can we say that I now see them as red?

The as is basically the idea of qualia. Check out the node to see what I mean.

What is really interesting about the inverse spectrum argument is that there are emprical examples of things very close to it. One experimenter gave people goggles that turned everything upside down. He asked them to wear them for a long period -- I think it was several months. After time, the people adapted. They could go about their lives, perform normal tasks -- some of them could even go skiing. But they just didn't have the vocabulary to explain the change that had taken place. They just became frustrated when people asked them: "So, do you still see everything upside down?"

I personally think that the inverse spectrum argument is based on assumptions about the nature of perception that simply don't work. These assumptions are brought to light very well by thinking about the inverse specturm argument -- which is why it is such a good intuition pump. One of these assumptions is the idea that there is a kind of perceptual "space," where things get arranged, a world in miniature that models the perceived world. This is a wishful assumption, I think.

BTW, there is a science of psychophysics -- I've taken a course in it. Scientists like Weber and Fechner (see Weber's law) made very rigorous scientific investigations into psychophysical laws, without missing a methodological beat. They asked people to report how the intensity of their experience varied with variations in objective stimuli.

To answer the question "Intensity of what" you'd have to ask the participants in the experiment. Weber simply varied something measurable (the amount of light, or sound, for instance) and correlated it with reported changes in the intensity of the subject's experience. That's science, not ontology. But ontology had better listen to it, to avoid ruling out things that have already happened. :)