From PhilPapers forum Epistemology:

2010-06-27
Does direct realism make sense?
There is a great silly song from before the War (that's 1939-45!) that says that when it's midnight in Italy it's Wednesday over here. It's Sunday over here and the direct realists haven't taken the bait yet so I am sending a message of sympathy. But the point of the silly song is that it presents a pseudodynamic idea. Sung well it has made me laugh to tears and I suspect laughter is our way of recognising a threat to our reason that proves empty. Messing with our concept of time is a serious threat, so when it is empty it is a big laugh. The response to magicians is different - the threat looks as if it is real - so we gasp. 
You identify that, like Free Will, direct realism is a pseudodynamic concept. I remember that I was quite convinced of it at age 21, yet by 30, after listening to a physiologist describing the genetics of taste, I had come to realise that I had been convinced of what seemed to be a dynamic concept but which cannot have been because there is no such concept. If you try to send it round the brain circuits that check that things add up (as Popper recommends) you find there is nothing to send. 

However, I often come across the statement that most philosophers are direct realists these days, so what is going on? Part of the problem is that the language we use can work with the same words in all sorts of metaphysical frameworks. This means what any explanation of direct realism means will depend on the metaphysical position of the reader. So if a direct realist replies we still will not know what he means unless the context is rich enough to give a clear idea of how all the words are being used. It may be that there are positions from which one could justify what is described as direct realism and yet agree with your position. So far I have the impression that there is no coherent position from which the concept can be rolled out and work, but I have an open mind.

As I understand it many philosophers returned to direct realism because it was thought that representationalism required the positing of some extra 'thing' called a mental representation or sense datum. However, since the whole concept of 'thing' as something used in an explanation now looks like a pseudodynamic fake, it would seem that if people want things then maybe the mental representation rescues them from a threat more terrible than the song. I am currently reading Rorty's attempt to foil every possible argument from every possible direction on this and abolish mental representations while still allowing people to have them if they find them cuddly. 

Maybe some direct realists will appear above the parapet tomorrow.

Best wishes

jo E