I read this because the title seemed offensive/wrong, or at least highly provocative, and I was curious to see if the author could make a reasonable and data-supported argument in favor of physiognomy. He made some reasonable arguments, but not conclusive. There were also a lot of editing failures in the book (bad formatting, inconsistent levels of detail). However, I'm going to be pretty charitable because it wasn't completely and obviously wrong while making a provocative assertion; it just wasn't definitive.

One of the immediate responses to this idea is that it's racist -- ironically that's one thing which he pretty clearly and convincingly argued against, as almost all of the correlations are only valid within a given race (and generally only within even more closely related groups, such as ethnicity or nation). The nazi abuse of phrenology was cited as being particularly abusive to data. There are reasonable gender-based arguments, supported by both data and biology, but a lot of the most interesting pieces were differences with 0.1-0.3 level correlation internal to gender/ethnic groups. ( )