Dear Glenn,But the subject in the recently defunct neuron that sensed that it was I (for those that still have this sense) will no longer have any sense of anything, so no sense of any lack of anything. Whatever subjects are now having the sense of being Glenn are clearly not those that died off a minute or two ago.

I think a neuron in a vat, suitably connected up, can be hypothesised to have an experience that is no meaningful sense different from that which a neuron has in your head at present. That would include the same content as defined extensionally and as far as possible intensionally as well. It might be very hard to ascertain this but it would be the working hypothesis that one would hope to survive any possible test.

As to what would determine if the content involved colours or sounds, that would be a very interesting question and presumably it would have to do with the interrelationships between the spatial location of the relevant post synaptic potentials. I am wary of calling content 'visual' or 'auditory' in this context because those terms refer to earlier sensory input pathways and in a vat you would not need to have separate pathways. The brain scientist studying the cell could connect up a programme to the cell that input signals giving colour and sound experience from the same electronic set up.

There is also an important pitfall here. For a cell capable of sustaining experiences of both colours and sounds there should be a set of rules for which input patterns were perceived as colours and which as sounds. However, there is no reason why another cell, with a slightly different computational role in a brain might use the same patterns for colour that the first used for sound or vice versa. A cell within the auditory signal collating system would use all patterns it had available for dealing with sounds, so it could make use of patterns that a multimodal cell might allocate to colour. There is no need for the same set of rules across all cells.

Taking that into consideration it is clearly very difficult to set up any very specific hypothesis about which patterns of post synaptic potentials are used in neurons for experiences of a particular extensional content. But I think we have to try to explain qualia with patterns of interrelation within individual integration events in individual cells. If you try to explain qualia by patterns of action potentials distributed across the brain you have absolutely no reason to think they would 'feel' any different from each other because they have no immediate causal relation that might encode different feels.