Even if this hadn't been quite such a good book as it is, I would have given it five stars for being neither about analogy nor pathology. I am tired of both, because as much as it is handy to refer to computer data storage as "memory," it really is nothing like human memory, and as much as my mother sees ghouls of Alzheimer's over every lost pen, the truth is that her memory isn't as good as, well, as she remembers it being.
Without being about pathology, this book is about the fallibility of memory; or rather I should say, the failure of memory to live up to the expectations that we have for it. Actually, this book has made me think about the purpose and function of memory, and I've concluded that it actually works rather well; if we had little videocameras in our frontal lobes, they wouldn't serve us as well as the memory functions we actually have, and in fact, this is the subject of the final chapter.
The seven "sins" of memory are transience, absent-mindedness, blocking, misattribution, suggestibility, bias and persistence.
Transience is the deterioration of memory over time, other than traumatic memory-- Persistence is the stubbornness of traumatic memory to fade. Absent-mindedness is failure to pay attention to something unusual that happens while performing a task by rote. Misattribution is attributing one feature of a memory to another-- remembering a childhood friend by his dog's name, for example. Blocking is the "tip of the tongue" phenomenon. Bias is coloring old memories with present knowledge.
There is no branch of study, from cranial anatomy, to neurochemistry, to performance psychology, to forensics, that he does not probe for usefulness. I applaud him for undertaking this project. In general, his writing is clear and concise. If occasionally he seems to belabor a point, this is something his editors should have corrected, and I don't take him to task for it. Skim through and go on.