Humans have a "triune brain": We have a reptilian brain, a limbic (mammalian) brain, and a neocortex, wrapped around each other in layers. The neocortex is the part of the brain that gives us reason, will, and presumably self-consciousness (animals like rabbits have very little neocortex, cats have a little more, monkeys more still, and humans the most of all). The limbic brain is what gives us emotionality. The reptile brain is what keeps us alive. Some illustrations: The limbic brain is what causes a cat to look after its kittens while an iguana walks away after laying its eggs, and what causes the cat to snuggle up to you and purr while an iguana would just as happily sun itself on a rock.
The limbic brain is capable of learning, just like the neocortex is, and this turns out to be a key finding. The limbic brain learns skills and feelings, while the neocortex learns facts. The book cites an example of a man who had a damaged hippocampus, so he was incapable of forming new (neocortical) memories, so that if you introduced yourself to him, he would ask you 30 seconds later who you are. But they experimented on him, and they found that (a) after teaching him to braid, they asked him if he knew how to braid and he said no, but when they placed three strips in his hand, he immediately braided them; (b) they instructed three nurse-aides to act according to the roles of "good guy", "neutral guy", and "mean guy", and after a few days they asked the man who he would prefer to ask for gum or a deck of cards, and he would consistently pick out the "good guy" even though he didn't consciously know any of them. (Interestingly, this lends credibility to the sub-plot in Dollhouse (that I used to think was ridiculous) where two of the dolls are in love with each other, even though they are kept in a state of neocortical blankness and don't know who each other are.)
So the first takeaway is that this is why we as humans can know something to be a fact, but still habitually act as though things are different. We can understand why something isn't working, and even be very motivated to change it, but none of that affects the limbic behavior that keeps on following its established grooves.
Now, the limbic brain is a biological advancement that comes with a drawback of requiring the existence other limbic brains to attune with. This is why mammals are social creatures (to a greater or lesser extent; there are mammals who are solitary as adults, but many species are wired to depend on social connections throughout life). Humans require meaningful relationships and interaction with other humans in order to stay emotionally healthy. A recent TED talk on longevity gave social support as the #1 most consistently present factor among populations with a high proportion of centenarians. There are even studies showing that people with pets tend to live longer.
Infants, particularly, require limbic bonding in order to develop properly. "A child is born with the hardware for limbic sensing, but to use it skillfully he needs a guide. ... A child makes constant use of his limbic link to adjust his impressions. The drama plays out a dozen times each summer afternoon at a local park. A toddler lurches across the grass with a determination that his unsteadiness renders positively quixotic. Inevitably gravity catches up with inexperience; he teeters and falls. At once he checks a parent's face: if she shows alarm or concern he cries, and if she is amused he may smile at her, even laugh. ... He looks to his mother as a piano tuner looks to the sound of pure C." In fact, a very young infant relies on constant communication: "Imagine a double video camera setup, in which mother and baby can see each other, but not face-to-face; each sees the other in their respective monitors. In real time, mother and infant look at each other, smile and laugh, and both are perfectly happy. If the baby sees a videotape of his mother's face instead of the real-time display, he quickly becomes distraught." A study of co-sleeping by James McKenna found that "[O]n a minute-to-minute basis, throughout the night, much sensory communication is occurring between [the mother and the baby]." In fact, the US's common practice of putting babies to sleep in separate rooms from the parents may account for our highest incidence of Sudden Infant Death Syndrome in the world - in some other countries, the syndrome is unheard-of.
If a baby mammal is separated from its mother, there are evolutionary responses that emerge in two stages. First, there is "protest," in which e.g. a puppy "paces tirelessly, scanning his surroundings from all vantage points, barking, scratching vainly at the floor. He makes energetic and abortive attempts at scaling the walls of his prison, tumbling into a heap with each failure. He lets out a piteous whine, high-pitched and grating." (There are parallels in human relationships: "The tormented letter that a rejected lover composes turns out to be an updated version of a baby rat's constant peep: the same song, in a slightly lower pitch.") Also, in this stage, physiological responses include the secretion of adrenaline and cortisol (stress hormone), presumably so that the animal can better find its missing mother - resulting in elevated heart rate and disrupted ability to eat and sleep. After some time, a mammal enters the second stage, "despair," in which he "stops his back-and-forthing, stops whimpering, and curls up in a despondent lump. He drinks little and may show no interest in food at all. If a peer or playmate is introduced into the pen, he may regard him with a bleary eye and turn away. He will have a slumped, dejected-looking posture and sad facial expression." There is a picture in the text of an "isolated rhesus monkey" which is heartbreaking: a little creature curled in the corner of a cage lying on its side in the fetal position, with arms and tail wrapped around its body, and wide sad eyes staring toward the ground at nothing. (A parallel for humans would be grieving the death of a loved one. Also, there is speculation that major depression may be a "twisted variant" of the despair reaction, and the similarity is close enough that "despair in laboratory animals is often used as a model for depressive illness.")
The long-term effect of insufficient or inconsistent limbic bonding in early childhood is a damaged nervous system, much like kittens raised with one eye covered grow up to be cats with "marked aberrancies in the brain areas serving vision." Monkeys raised in isolation exhibit behavior like self-mutilation, unpredictable violence, prolonged food and water binges, timidness, clinginess, and clumsiness in efforts to establish ties to other monkeys. They are also much more sensitive to anxiety and depression, the adult analogues to protest and despair - these responses can be evoked by relatively minor incidents and can persist for longer. In humans, the result can be a twin problem of increased need for human contact, but increased difficulty in making it. "Without rich limbic resonance, a child doesn't discover ... how to tune in to the emotional channel and apprehend himself and others. Without sufficient opportunity for limbic regulation, he cannot internalize emotional balance. Children who are thus handicapped grow up to become fragile adults who remain uncertain of their own identities, cannot modulate their emotions, and fall prey to internal chaos when stress threatens. ... A person who lacks a stable center feels an urgent need to fill the gap -- he needs *something* to orient himself as he tries to navigate the world." This was the second useful takeaway for me.
The third interesting topic in the book was the way humans select relationship partners. The authors compare the limbic brain to a neural network (an AI technique), and explains that we learn early on what healthy relationships look like by observing our parents, much as a neural network might be trained. We then have "attractors" (patterns) that are constantly scanning for a match, someone who will interact with us according to our internalized template. This can explain why a man who had a critical mother may have a recurring pattern of dating women who start out wonderful but turn into cranky imitations of his mother. It is both because he seeks out women with this potential, and also because once in the relationship, he unconsciously engages in behavior to stimulate her to act this way, in order to fulfill the template.
The book does also talk about ways to fix broken adult limbic imprinting, and devotes a whole chapter to client-psychotherapist relationships and their potential for success. But the message is not terribly upbeat: it is possible to change and heal limbic imprinting, but it takes years of persistent work with a capable partner since adult circuitry is not as malleable as that of a child. The real advice of the book seems to be: if you have a child, for heaven's sake pay enough attention to it. :)