Scientist Mary and Causal Theories of Reference

I want to draw some connection between the Jackson’s Knowledge Argument and the causal theory of reference. I will probably say lot of problematic things on which people don’t agree, without saying that those things are problematic. That isn’t because I’m sure those things are as I say they are, but just so that side comments don’t obscure the relation I want to draw. So here it goes…

To be red is to appear somehow in specific circumstances. Let’s leave aside what are those specific circumstances. My inclination is to talk about “uncomplicated” circumstances, but maybe it should be ‘normal’ or ‘ideal’ or ‘optimal’ or something else. People became aware that some things have some specific appearance which they also could remember and recognize, and used ‘red’ to refer to it.

I can’t say that “to be red is to appear red (in specific circumstances)”, because I take it that to say that something appears red (in some situation) is like saying that the thing appears same as red things appear (in specific circumstances). And so that would render “to be red is to appear red” circular.

Scientist Mary knew about red color (e.g. that there is some color which is referred by the word ‘red’), knew how to recognize red things (using technology for example – a red things detector) and so on, but she is not acquainted with red things’ appearance in terms of their color. She might have seen red things, but she never have seen their color (say, red things were presented to her, but because of some operation on her eyes she was temporarily fully color blind). What she learns then when she lives the room is how red things appear (in uncomplicated circumstances). But Mary doesn’t learn just that. Because she knows that red things in uncomplicated circumstances appear same as white things appear when shined by red light, she has also learned how white things appear when shined by red light.

But one can do the things the other way. By presenting Mary with a white ball shined by red light, she can learn what white ball shined by red light looks like. But as she knows that white ball shined by red light appears as a red ball in uncomplicated circumstances appear, she now has learned what red things in uncomplicated circumstances look like.

But if to be a red thing is nothing more than to appear somehow in uncomplicated circumstances, there is nothing more to learn about what ‘red’ refers to than what Mary became aware by seeing a white ball under red light. Or maybe red things don’t enter the story anyway, even red lights. Maybe Mary was presented with a green circle and then was asked to look at a white wall. The wall because of the afterimage illusion will appear same as a wall with a red circle on it. So Mary can become aware of red, being presented with situations in which there are no red things nor red anything.

Let’s change the scenario a little, and say that people were hiding the names of colors from Mary. After seeing the red afterimage, Mary can form idea of things which appear in uncomplicated circumstances as the wall appears with the afterimage effect, name the color of those things ‘red’, and ask ‘are there things with red color?’. So, now Mary has a name for red color (a property that red things have) without ever being acquainted with things with red color (nor anything red). (Of course, she might not call it ‘red’, but the fact is that she has word for red, without ever being causally related with anything red, nor is the meaning of the word borrowed meaning.)

What if she didn’t know about afterimage illusion, so that she wasn’t aware that she is seeing just a wall in “complicated” circumstances. As in the previous case, she is aware of everything that one can be aware of about red, can continue using ‘red’ to refer to red things, and might in fact after some time come to know, that what she saw the first time was not a red thing, even she did the baptizing on base of something that was not red, nor was causally related to anything red. She can say “I thought it was red thing, but it was just an afterimage”.

Is this scenario compatible with causal theories of reference?