22. The knowledge argument

Martín Abreu Zavaleta

July 1st, 2014

1 The argument

Jackson’s argument can be reconstructed like this:

At time t0, Mary has all the physical information concerning human color vision.


Jackson thinks that qualia, or phenomenal properties of mental states are causally inefficacious: their possession or absence makes no difference to the physical world. However, he thinks that the instantiation of qualia may make a difference to other mental states. Most of his defense will consist in rejecting some reasons to think that qualia are causally efficacious.

It seems obvious that the way pain feels is partly responsible for our reactions to pain, so the way pain feels must have some causal powers. Jackson merely points out that just because two phenomena are correlated, that doesn’t mean that one causes the other. In particular, if there is some more plausible theory (e.g. one that explains how the two phenomena are caused by a third thing, but independent of each other), we should reject that the way pain feels has any causal powers.
Some people claim that we may assume that qualia evolved over time, and so we may expect qualia to be conducive to survival. After all, according to natural selection the traits that evolve over time are the ones that are conducive to survival. Jackson points out that not every trait has to be selected. It may be that some trait is not conducive to survival, but is a by-product of something that conduces to survival. If this is the case, evolutionary theory is compatible with epiphenomenalism.
Our way of knowing that other people have qualia depends on our observation of their behavior. If qualia were not causally efficacious, observation of people’s behavior wouldn’t give us much reason to think that they have qualia. So qualia must be causally efficacious. Jackson continues his strategy and points out that behavior may still give us evidence for the presence of qualia in other people, if only because we may postulate that whatever causes the behavior causes the qualia too.
If qualia weren’t causally efficacious, they would play no explanatory role. Our best theory would not need to posit them. Jackson acknowledges that qualia may not play any interesting explanatory role, but claims that this doesn’t show that there are no qualia.

2 Refinements and observations

The first thing to note is that, the way Jackson originally stated the argument, there is an ambiguity in the notion of information. In one sense, ‘Hesperus is a planet’ and ‘Phosphorus is a planet’ transmit different information, since someone may know that Hesperus is a planet but still not know that Phosphorus is a planet. In the other sense, ‘Hesperus is a planet’ and ‘Phosphorus is a planet’ give us the same information, since they express the same proposition, or are true under exactly the same circumstances. Which sense of information is the one required by the argument? Let’s examine both.

For the sake of clarity, let’s call the first notion of information narrow and the second wide.

With narrow information, we may present the argument as follows:

At time t0, for every truth p about color vision described using only physical vocabulary, Mary knew that p.
Before Mary comes out of her lab, there is some qp such that Mary didn’t know that q.
Therefore, not all knowledge is physical knowledge.

Here the conclusion is to be understood in the following terms: that one can know something when described in a particular way, but not under a different way. This is just like knowing that Hesperus is a planet but not knowing that Phosphorus is a planet.

This sort of conclusion need not worry the physicalist. Physicalism is usually meant as a metaphysical thesis about the nature of the mental. In its simpler formulation, it claims that mental things are physical things, or reducible to physical things. This metaphysical claim is compatible with the claim that mental vocabulary may just not be translatable to physical vocabulary, or that mental concepts are not reducible to physical concepts.

However, there is a stronger version of the argument, using the wide notion of information. Jackson’s own presentation is as follows:

Mary (before her release) knows everything physical there is to know about other people.
Mary (before her release) does not know everything there is to know about other people (because she learns something about them on her release); therefore
There are truths about other people (and herself) which escape the physicalist story.

On this way of understanding the argument, what Mary doesn’t know is a fact. It’s not only that she lacks the appropriate concepts, or that she doesn’t know under a particular description. What she fails to know is a fact, and given that she knew all the physical facts, this new fact can’t be physical. Questions: What do you think of this version of the argument? Has Jackson made a compelling case that the main premise is true?