Book review of Edmund Rolls

Edmund Rolls:
THE BRAIN AND EMOTION (Oxford Univ Press, 1999)

Home | The whole bibliography | My book on Consciousness

This book provides a very intriguing and very comprehensive survey of brain processes related to emotion, related by a very competent scholar (an Oxford psychologist), who advances the hypothesis that the fundamental design of the brain is built around the mechanisms of reward and punishment.
The brain determines which action to carry out based on such reward and punishment mechanism. The brain is capable of computing the "value" (in terms of reward and punishment) of a sensory input. That capability comes from millennia of natural selection. Somehow this process of reward and punishment expresses itself through emotion and motivation: the brain computes the "value" of a sensory input through the brain mechanisms of emotion and motivation. For example, when the body needs food, the brain generates the emotion of hunger. That motivates the body to eat. When the body has eaten enough, the brain generates the emotion of satiety.
The brain is designed around the reward and punishment mechanism. The brain plans behavior that helps obtain rewards and avoid punishments, i.e. "motivated" behavior.
The book then delves into the technical details, the experimental findings, about the regions of the brain that are responsible for motivated behavior, in particular, those that provide orientation to environmental stimuli.
Rolls first deals with emotions as "states produced by reinforcing stimuli" (a purely physical, neural approach) and only later with the "feeling" of an emotion. His focus on "reinforcing" stimuli helps solve the puzzle of why only "some" stimuli elicit emotions. The ones that are "reinforcing" need to be decoded by special regions of the brain, which generate the neural processes that we call "emotions".
Rolls' book is mainly a very detailed discussion of the neural circuits that produce emotion and that produce behavior in response to the events that have caused the emotion.
Rolls shows that there are different circuits for primary reinforcers (pain, touch and the smell of food) and secondary, or learned, reinforcers (the ones that become reinforced by association with a primary reinforcer, such as the sight of favorite food). Impressive progress has been made in understanding how the amygdala and regions of the cortex are involved in learning reinforcers and in decoding the known reinforcers.
The neurophysiology of reward (and the addictive role of dopamine) is particularly interesting, as it may have practical applications in the short term.
Emotions have several functions, including: the production of an autonomic response; (i.e., faster heart beats); the production of an endocrine (i.e., adrenaline) response; the production of motivated behavior (not only specific for the event at hand but even a direction of behavior that will last a lifetime); a re-evaluation of the stimuli themselves for future use; communication to other members of the group; the storage of "important" memories; etc. Rolls lists ten functions overall.
The representation of a "reinforcer" (i.e., an object that causes an emotion) contains information about how to calculate reward/punishment. Not surprisingly, the representation lends itself to generalization (if one tiger causes fear, all tigers should) and association. The emotional aspect of a perception is basically due to the activation of its representation.
The brain is designed around the mechanisms of reward and punishment. Whatever system is in charge of "action", that system must be taking the output of reward/punishment assessment as its goal: the brain as a whole wants to maximize rewards and minimize punishments, i.e. to maximize the activation of representations relative to rewarders and minimize activation of the representations relative to punishers. The specific behavior is not genetically fixed: what is fixed genetically is the goal, which is to maximize reward and minimize punishment.
The chapter on sexual behavior links sociobiology with brain physiology. The brain has evolved to produce sexual behavior that increases biological fitness. The ultimate purpose of sex is successful reproductive behavior: passing one's genes on to the next generation. Rolls argues that natural selection favored not one but a whole range of reward systems, and that accounts for the variety of sexual behavior that we encounter among humans. Thus some men and women are genetically predisposed to be unfaithful.
What sets apart sexual behavior from other reward systems (e.g., hunger and thirst) is that there is no homeostatic control: while hunger and thirst turn themselves off when an appropriate level of feeding and drinking is reached, sexual appetite continues. Rolls explains that sexual behavior is influenced by sperm-level competition among males to impregnate females, and mate selection by femals to choose the best male.
Only towards the end does Rolls turn to the "feeling" (the qualia) of emotions, and, in doing so, he ends up proposing a new theory of consciousness. There are two routes to action that can be followed in relation to reward. One route is shared by humans and other primates, so it is probably the primitive one, and centers on the amygdala. This route is a "shortcut" that sends stimuli directly to the motor system of the body. A second route involves "planning" the action, and therefore "thinking" about it, and centers on the language system. The second route is what allows us to make "long-term" decisions rather than only short-term decisions. It is also the one that Rolls presumes is causing consciousness. Consciousness is therefore a property of the brain processing rewards, a side effect of high-level planning.
Rolls implies that consciousness arises from the capability of thinking about one's thoughts, or higher-level thinking. He goes on to suggest that this feature could have an evolutionary value, as it allows to correct mistakes of first-order thinking (in other words, to learn). Language is therefore necessary but not enough for consciousness: consciousness also requires the ability to think about thought, or higher-level thought. That said, some sensory input become conscious (become "feelings") precisely because this way they can be analyzed by higher-level thought. In a sense, the mechanism of higher-level thought would be pretty useless if it couldn't use the feelings associated with sensory input.
Patterns of neuronal firing in the brain correspond to patterns of stimuli in the world, which is to "objects". Therefore there exists a representation of an object in the brain, and such representation is the corresponding firing of neurons. Such a mental representation obtains its meaning in two ways: from the associated reward (or punishment) "value" and from its sensory-motor correspondence in the world. First, some actions are but the means to activate such representations, because living beings are programmed to achieve reward and avoid punishment, and their search for rewards (and away from punishments) is basically simply a search for "activation" of those representations. A brain will work to make those representations become active. Secondly, each "object" is associated with the kind of actions that are possible and not possible with it. The same arguments apply to language: the word "banana" has meaning both because its taste has a reward value and because it is associated to the action of peeling.
Although it deals with topics that could interest a broad mix of readers, this is not an easy book to read because the style of the author is rather convoluted. To quote one of the key conclusions, "animals must be built to want to activate the neurons in this system which can be activated by stimuli learned to be reinforcers in the amygdala..." It takes a while to appreciate the breadth and depth of Rolls' claims.

Permission is granted to download/print out/redistribute this file provided it is unaltered, including credits.