Shared publicly  - 
Aggression II. 105m Robert Sapolsky video.

Part 2 of 4 in a broad biological survey on the biology of morality, empathy and aggression. After completing the discussion of the amygdala, Sapolsky turns his attention to the frontal cortex, the part of the brain which is larger proportionally in humans than in any other species. He goes in depth on the behavior and relationships between the frontal cortex and other brain regions (including its special relationship with the amygdala). Then he looks at the interconnectedness of the cortical and limbic functions: the supposed separation between the rational and emotional brain has crumbled with modern science. Then he discusses the fascinating topic of metaphor in the brain. He ends by starting the discussion on the hormonal effects of aggression. Extensive notes below.

18. Aggression II

People with amygdaloid damage cannot detect fear-evoking faces, they're overly trustful, underly skeptical, they miss cues that the situation requires arousal or vigilance, and they do not look at the eyes as much as normal individuals (Antonio Damasio work). Not only is the amygdala evaluating if this situation demands fear but it is on the lookout for it. Testosterone makes the amygdala better at detecting fear in anger-invoking faces. The lateral geniculate nucleus, the primary relay center for visual information, has a shortcut to the amygdala (which may be extra excitable in post traumatic stress disorder PTSD). The shortcut trades off speed in recognizing danger for analytical accuracy. The amygdala is one synapse from olfaction. It is connected to the frontal cortex.

amygdala = fear = anxiety = aggression

Exceptions: Williams syndrome is a poorly understood imprinted genetic disorder resulting in kids who are unbelievably facile with language and emotional expressivity yet borderline retarded cognitively. Very affectionate, trustful, gregarious and so incredibly vulnerable to being taken advantage of. They do not react to scary faces in their amygdala. People with social phobias have the amygdala activate for any face: for them every face is a scary face. In depressed people the amygdala activates when you show them something sad suggesting that the amygdala is subtle and contextual in its functioning.

The amygdala is probably the best part of the brain for dichotomizing between us or them and responding to outgroup stimuli. The amygdala forms these dichotomies very quickly.

Going in depth on the Frontal Cortex

Frontal cortex is involved in regulating appropriate behavior in the context of violence, aggression, competition, & cooperation. It was Walle Nauta who suggested that the wiring of the prefrontal cortex makes it part of the limbic system (the so-called emotional brain). When the choice is between something harder (yet better for you) and something easier, the frontal cortex is about doing the harder thing. The frontal cortex has many weak, diffuse projections throughout the limbic system to give more biasing to excite those pathways that are "harder". The frontal cortex provides a modulatory function. Dopamine fuels expectation and drives the frontal cortex to goal-directed behavior.

CVLT (California Verbal Learning Test) memory test: repeating a list of things purchased. By fourth round one starts grouping the items into categories ("executive function"). The frontal cortex helps us find a better strategy (by categories): identifying the pattern in the data. People with frontal cortex damage never start grouping the items. Frontotemporal dementia: drawing the clock hands at 11:10 (10m after 11): drawn at 11:50 (11 hand & 10 hand): they are drawn by the easiest interpretation of the numbers "11" "10".

The frontal cortex can inhibit a well-learned task. So with frontal damage repeating the months in reverse order only works for a bit before the habitual behavior takes over. "Intrusions": count from the number 20 and slip back to giving the months in the usual order. Verbal fluency test: in 1 minute say as many words beginning with the letter 'f', then repeat with different letters. Those with frontal damage go back to the task with the letter 'f' again.

The frontal cortex is good for making you work for a cognitive reward way down the line such as working at college so you can get into the nursing home of your choice in sixty years. Gratification postponement: doing the harder thing.

In eltrophysiological studies (electrodes in the brain for recording excitations) with recording electrodes in the visual cortex and the frontal cortex (in a monkey), the frontal cortex remembers the rules of the game and so stays activated once the signal that the game has begun is sensed. The frontal cortex does not code for individual examples of rules but maintains rules over time. This is hard work and they are subject to failure and require lots of energy. When something is learned and becomes automatic or reflexive, the frontal cortex stops being activated so the learning is stored elsewhere in the brain (the cerebellum).

Phineas Gage had a large iron rod destroy his frontal cortex. Even though he was able to walk to the doctor, his personality changed from foreman to a brawling, abusive, sexually predatory, out of control individual. The original doctor deduced that whatever part of the brain that is, it reins in our animal energies. More recently an 80 year old man who had a stoke damage his frontal cortex, raped an 80 year old woman with Alzheimer's disease. About 25% of men on death row in the US have a history of concussive trauma to their foreheads.

Legal implications of brain damage and criminal behavior. M'Naghten rules: law on the books of most states (and several other countries) for organically impaired defense for criminal liability (so-called insanity defense): can the individual tell the difference between right and wrong (no evidence of them trying to cover their tracks: no apprehension of having done anything wrong). But some people with frontal cortical damage can tell the difference between right and wrong: they can state the rules, but cannot control their behaviors. M&M test: show a hand with 5 M&Ms and another with 1 M&M: if the subject reaches for the hand with 5, give them 1 M&M; if they reach for the hand with 1, give them 5 M&Ms. People with cortical damage can learn the rule, but always reach for the hand with 5 M&Ms (the easier path: no executive perspective). Organic impairment of being able to follow the rules; not organic impairment of knowing the rules! In the aftermath of John Hinckley's successful insanity defense, the federal government revoked the M'Naghten rule in federal criminal trials and most states have repealed M'Naghten as well as volitional impairment rules.

Chimps with less frontal cortex cannot pass the M&M test. But if wood chips are substituted for M&Ms (removing the temptation), then chimps pass the test. Children also are unable to pass the M&M test: marshmellow test: here is a marshmellow, if you haven't eaten it before I come back then you will get two. How long kids can hold out is predicted by the amount of frontal cortex metabolism rate (activity) they have. The length of time a five-year old can hold out predicts SAT scores. Hide and seek test: After countdown say "Where are you" and most kids will instantly say "right here" because they don't have enough frontal cortex function to restrain themselves. When the kid counts down, they can't inhibit themselves from stopping at 10.

One kid had his frontal cortex damaged in a car accident at age 6, he committed his first murder by age 13 and became a serial murderer. One of the women he assaulted, he drove home and gave his number and asked her to give a call saying he had a really good time. But he could recognize that it was wrong if his case was given with other names (he even judged the case to be criminal). If the frontal damage comes before age 5 or 6, they never learn the rules: acquired sociopathy. After 5 or 6 there is a partial learning of the rules with the inability to act on it. Adult damage results in a knowledge of the rules but inability to restrain the behavior.

Another case of a child with frontal cortical damage from foreceps slipping at birth. Frontally disinhibited: unconstrained by society or its laws. But in this case its an upper middle class family with excellent family support, but he plays the piano way longer that is socially acceptable. So environmental differences can result in the damage manifesting in different ways.

So some states protect people (M'Naghten rules) with 99% of their frontal cortex destroyed. What about 97% or 85% or 45% damage? Or you and me who have 5 more or fewer synapses than the other in our frontal cortexes?

Epileptic seizures in the 16th century were medically explained as demonic possession punished by burning at the stake which was proven by observing if they cried when read the story of the crucifixion. Someone at the time observed that in some elderly women the lacrimal gland can atrophy making them unable to cry.

During REM sleep, the frontal cortex is almost entirely inactive which is why dreams make no sense.

The frontal cortex is the last part of the brain to be fully myelinated (the white sheath that speeds neuronal communication) and to develop its branching synapses. It is finally fully developed about age 25 hence it is the part of the brain most sculpted by environment and experience and so least constrained by genes. "It is the most definably human part of the brain."

The dopamine-driven metabolic changes (rewards and disappointments triggering rises and falls in dopamine) swing much more widely in teenagers than in adults. The teenager's brain is less regulated than the adult brain.

Supreme court decided that since the brain and its regulatory functions are not fully mature, the death penalty should not apply to kids under 18. But brain science offers no reason to assume that one's 18th birthday resolves any of the frontal cortex deficiencies of one's late teens.

The frontal cortex is the third most vulnerable brain region to aging. The substantia nigra a motor system in the brain which loses most of its neurons with aging which causes the tremors of old age and Parkinson's disease. The hippocampus loses most of its neurons with age which affects memory in the elderly. The degradation in the frontal cortex with aging explains disinhibited comments from the elderly.

People with repressive personalities (highly regimented, disciplined, controlled behavior, weak at expressing and reading emotions in others, not depressed nor anxious) have elevated resting metabolism in the frontal cortex. Sociopaths have low activation (low metabolism rates) of their frontal cortex; when they recite the months backward (requiring frontal cortex function), it takes them more work to do it than for those whose frontal cortex is generally more active.

By age five, the size and resting metabolic rate of the frontal cortex shows the substantial impact of socioeconomic status (poverty) on glucocorticoid levels (from the stress of poverty; the frontal cortex has a lot of glucocorticoid receptors which atrophy neurons with elevated levels) which inhibits neuronal growth.

It is not clear why frontal cortical damage results in disinhibition leading to murder rather than disinhibited generosity or love. But evidently the frontal cortex is more important in regulating the amygdala. The amygdala and frontal cortex have a bidirectional inhibitory relationship: each inhibits the other: the frontal cortex is trying to get the amygdala to restrain itself, the amygdala is trying to get the frontal cortex to stop sermonizing at it. There is an inverse relationship between the metabolic level of the amygdala and frontal cortex under resting conditions in rats, primates, and humans. For example, in habituation to a fear conditioned response (extinguishing the behavior), the amygdala activation decreases as the releasing stimulus is learned. If the frontal cortex is damaged, the amygdala never learns to modulate its activation.

There are situations when both the amygdala and frontal cortex activate: whenever doing the harder thing is also the scarier thing such as trench warfare in WWI.

The septum is another brain region that modulates the amygdala. The septum inhibits aggression. Remember the rule for the limbic system: every brain region (nucleus) tries to enhance its signaling and inhibit the signaling of other regions to get the hypothalamus to "listen" more to it.

Lateral hypothalamus actually has nothing to do with aggression. Instead it is involved in predatory behavior.

Anterior cingulate, behind the frontal cortex, activates with pain perception even if it is someone else's pain: "where literally and not just metaphorically you feel the pain of other individuals". Josh Greene at Harvard uses functional MRI to study subjects asked about the hypothetical situation of hiding from the Nazis with a crying baby: is it ethical to kill the baby? Subjects who activate the anterior cingulate are less likely to consider it OK to smother the child (less activation and more subjects consider the smothering justifiable).

All this research shows that the old dichotomy about thought (cortex) and emotion (limbic system) is crumbling. Demasio's book "Descartes Error": in terms of brain function, thought and emotion are inseparable. In many realms it is a ridiculous dichotomy to separate pure abstract cognitive decisions and the messy emotive system.

Three cases where cortical function can be separated from limbic function

1. Josh Greene: runaway trolley problem: it is heading to kill five people. Subjects are given the choice to pull a lever to divert the trolley to a track where it will kill one person or to push someone in front of the trolley to stop it. People are 75% more likely to pull the lever (with cortical activation) than the more visceral pushing someone scenario (25% with limbic action). People with cortical damage are more likely to say it is OK to push someone onto the tracks in the runaway trolley problem (more utilitarian decisions).

2. Oliver Goodenough used brain scanning to determine that cortical activity was higher when a judge says to interpret this case according to the law instead of asking jurors to remember that the system is about protecting the weak (activation of limbic system).

3. Transmagnetic stimulation allows researchers to decrease the activity of a brain region for a few minutes at a time. Testing shows that using this technique people playing games become more utilitarian and selfish when you turn off their frontal cortex.

Metaphor in the Brain

Because our brain evolved to handle the abstract judgements of modern society (ruining someone's reputation or plagarism), some of these abstractions are "stored" in "old" circuits. So we treat the metaphorical components as if they are real. For example, the temperature detection circuits (hot & cold) also affect our rating of another's personality: experiment begins in the elevator where a confederate asks the subject to hold a cup of something warm or cold: later they judge the personality of the other person as warmer or cooler depending on the temperature of the cup they held in the elevator.

In the anterior cingulate the brain both senses pain and empathically senses the pain of another. Another metaphor / reality mixing.

Insular cortex activates when foul odors are sensed in many species. In humans, tell someone a story about someone being mistreated by someone powerful or being exploited (such as "stab you in the back" in a game) and the insular cortex activates. Moral disgust and culinary disgust co-occur in the same brain region. Every human culture has language referring to moral failures with words denoting gustatory repellent stimuli: "I'm disgusted by what you did", "it makes me nauseous", "something about this smells rotten". In another study, if people are asked to talk about moral failings, they are more likely to choose soap as a reward than other equally valuable commodities. We shoehorned our sense of moral disgust into the part of the brain that handles bad food intermixing metaphor and reality.

Jonathan Haight shows how much of human decision-making is affective: your limbic system decides before you do. The cognitive system struggles to explain what the limbic system has decided for you.

When someone does something aggressive, the sympathetic nervous system activates. You cannot tell the difference between murder and orgasm when looking at the basic brain systems. There is a lot of non-specific activation in arousal. Love and hate are physiologically very, very similar: arousal of the sympathetic nervous system. Elie Wiesel: "The opposite of love is not hate, it's indifference". Many humans can psychopathologically confuse love and hate.

Hormones of Aggression

Why are males generally more aggressive than females? Is testosterone the cause? Testosterone plays the same role in aggression as it does in human sexual behavior. Testosterone is required for the full expression of aggressive behavior, but even after castration aggressive behavior does not go to zero (prior experience also influences the result). Whether adding 10%, 100% or 200% of normal testosterone levels back after castration, aggression returns to normal levels. Testosterone is needed but is neither necessary nor sufficient for aggression. You cannot predict aggressiveness based on testosterone levels. In Rhesus monkeys with a dominance hierarchy 1>2>3>4>5. Pump #3 with testosterone and he becomes more aggressive, but only to #4 and #5. Testosterone exaggerates (modulates or amplifies) the preexisting social structure. Raise testosterone and the amygdala has a lower threshold for activating given a potentially threatening face and at the cellular level it shortens the refractory period after a neuron fires (so that it can fire again more quickly): again, testosterone amplifies the activity.

Spotted hyenas have female dominance where females are larger, more aggressive, and have higher testosterone levels than males. Female hyenas have an enlarged clitoris the size of a male penis and compacted fat cells that look like a scrotum so you cannot tell males from females. An interesting case of a sex reversal: androgenized females. Hyena feeding order is cubs then females then males (in lions it is males then females then cubs). Masculination of the female genitilia. Generally male dominance displays involve waving their erect penis about. In hyenas, males get erections when they are terrified as a subordination gesture (similarly, low ranking females get clitoral erections). Hyena females are hormonally more male than male hyenas to achieve the sex role reversal in their social structure.
Mark Arnold's profile photoLazăr Lung's profile photoCJ Fearnley's profile photo
Maybe I missed it in Aggression I... Is there a point to this?
Mark, these notes are my attempt to summarize Sapolsky's lecture. The point of the lecture is to detail what is known about the neuroendocrinology of morality, aggression, and empathy. I find Sapolsky fascinating and so I wanted detailed notes that I can search in the future to "remember" which lecture covered which details.
+CJ Fearnley here's a question, perhaps I missed the point: if chimps get 6 M&M's anyway, why should they delay gratification? What's the advantage in going for the one M&M first?
If they reach for the hand with more they receive fewer M&Ms. The experiment shows that your frontal cortex provides gratification postponement preventing you from the decision that looks better (there are more M&Ms in that hand) but will result in less reward in the long run (the rule is you only get 1 M&M if you reach for the hand with 5 M&Ms). I hope that helps. Watch the video to see Sapolsky explain it. My notes are abridged.
Add a comment...