Let me give you an overview of the first two chapters (the only ones I've read so far....) and the introduction of the book.
In the introduction, he promises to give an overview of the brain (or kludge) and to highlight in which way the brain is not perfect. His point is that the brain is not a beautiful machinery but that it is a kludge that has evolved over the years (millions of them) and that the engineers have never had the opportunity to start from scratch. Therefore, these engineers had to take into account the old design and to improve it.
In chapter 1, he gives an overview of the basic anatomy of the brain. I particularly enjoyed its description of the role of the cerebellum. Probably because that's one area of the brain that is of particular importance in motor control and motor learning in particular. In this chapter, Linden explains that the cerebellum plays a role in predicting the consequences of our own actions, which is why you cannot tickle yourself and why you always think that someone has exerted a higher force on you arms than you did on his. Unbeknown to me, he also highlights the existence of ancient visual areas (in the mindbrain) on which we do not usually rely (but are still there). They appear to be useful if you have a lesion in the usual visual areas. When the position of an object arises from this area only, you do not consciously perceive that object although you can reach and pick it up. Finally, he illustrates the role of the hyppocampal complex (memory) and frontal areas (mood, personhood) with the patients H.M. and Phineas Gage. Note that Henry Molaison's brain (H.M.) has been recently dissected and its slices will be digitized for further study by the Brain Observatory at UCSD.
In chapter 2, Linden describes how neurons fires and transmit information. This is clearly not what I'm most familiar with so I learned a lot of things. Off course, he particularly insists on the slow and imperfect mechanisms that have arisen from evolution. GABA, NMDA, poisonous fishes... They are all there to explain the very basic of Neurosciences: how does a neuron fires and how does it transmit information. He also comments on how well the brain can work even if each neuron is very limited. That made me think of the super-computers built with hundreds of Playstation in parallel where all units are rather limited compared to the best available computes but are , together, able to produce a huge power (http://www.ps3cluster.umassd.edu or juste type "playstation supercomputer" in Google). This is only an analogy but I thought it was worth mentioning. Also, one advantage of this very redundant network is that small lesions are not able to heavily disrupts how the brain functions. Only massive lesions have drastic effect (like in H.M. case). I'd better had my life depend on hundreds of playstation than on one single computer running Windows98 (which was really unstable....).
More thoughts on the "Accidental Mind" when I'll have some more time to read....