With all the flurry of articles about brain-scans - fMRIs, there's a great article about the fallacy of overintrepreting the lit areas of a scan. Your brain isn't made up of nodes or specific locations for specific activities. It's not like your brain has a spot for looking at a painting, or making a deodorant buying decision...even though research is testing those things, and looking for what 'lights-up' when you engage in such an activity. Scientific American - Five ways brains scans mislead us
(you can read the whole article here, sorry Scientific American, you charge $7.95 to get the full article and I found a pdf for free for you...it's worth a look)
I contend it's still very early days, and the fascination of fMRI testing is in full force. My favorite analogy of all I've read is that memory is more the bits in a blender -- the picture of that is so compelling and vivid. In this article Michael Shermer he makes a clear case for why the brain is more akin to a network, and offers good skepticism about applying too much pop-science to brain scans.