> Why would materialism/physicalism require that first hand knowledge not exist? If consciousness is explainable in terms of physical processes, then first hand knowledge would exist and be part of that process.

You cannot describe first-hand facts using only third-hand facts. This is the core philosophical dilemma in the hard problem of consciousness. Consider something like Mary's Room. What sort of third-hand description of "what it is like to see red" would describe the first-hand experience of seeing red? How can you actually capture data describing "what it is like"?

The only solution seems to be to deny that "what it is like" is first-hand knowledge at all, and that it is actually a set of third-hand facts that merely yields the illusion of first-hand knowledge, and so the hard problem reduces to explaining how this illusion comes about.

My question was, why is "denying first hand knowledge" the only solution? As you might know, there are several lines of thinking. Nagle, for example, might say that first hand knowledge absolutely exists, but it is "off limits" to anyone but the organism experiencing it. In that sense it exists and does not require us to say that it is only "illusions of third hand knowledge". On the other hand Dennett might say that once we have a full scientific understanding (as Mary would), we could in theory understand how to extract the first hand knowledge in an organism's brain and either modify our own brains to experience it or describe it so fully that Mary would know exactly what to expect when seeing red for the first time. I'm not sympathetic to Dennet, but there are other options than first hand knowledge simply not existing within materialism.

> Nagle, for example, might say that first hand knowledge absolutely exists, but it is "off limits" to anyone but the organism experiencing it.

If physicalism is true, then all facts must be reducible to the physical.

If everything is reducible to the physical, then the experiences of the organism are also so reducible and must have a third-hand description.

If experiences are not so reducible, then physicalism is false and no amount of third-hand knowledge can produce first-hand knowledge.

These are exhaustive and mutually exclusive options.

> I'm not sympathetic to Dennet, but there are other options than first hand knowledge simply not existing within materialism.

The meaning of first-hand knowledge in physicalism is different from the meaning of first-hand knowledge in other metaphysics. In physicalism, first-hand knowledge must be reducible to third-hand knowledge, and so it doesn't have a privileged ontological status, just like we don't ontologically commit to the existence of cars in quantum physics.

So first-hand knowledge cannot really exist in physicalism except as a useful label describing a specific kind of third-hand knowledge.