The Emotion Machine

ideas.

In chapter §8 we also reviewed various ways to represent knowledge and skills, and noted that these could be arranged in a stack with increasing degrees of expressiveness.

Each of those ways to envision a mind has different kinds of virtues and faults, so it would make little sense to ask which model is best. Instead, one needs to develop Critics that learn good ways to choose when and how to switch among them.

Is a Mind like a Human Community?

Perhaps a more popular model of mind portrays our mental processes as like a human community—such as a residential town or a company.[203] In a typical corporate organization, the human resources are (formally) organized in accord with some hierarchical plan.

We tend to invent such ‘management ‘trees’ whenever there’s more than one person can do; then the work is divided into parts, which then are assigned to subordinates. This picture suggests that one might try to identify a person’s Self with the Chief Executive of that company, who controls the rest through a chain of command in which instructions tend to branch down from the top.

However, this is not a good model for human brains because an employee of a company is a person who might be able to learn to perform virtually any new task—whereas, most parts of a brain are too specialized to do such things.[204] This difference may be important because, when a Company becomes wealthy enough, it may be able to expand itself: when it wants to include new activities, it can hire additional employee-minds. In contrast, we humans don’t (yet) have practical ways to expand our own individual brains. So, whenever you add a additional task (or break a large one into smaller parts)—and then try to do them all at once—then each sub-process will lose some of its competence, because it now has access to fewer resources. Perhaps we should state this as a general principle.

The Parallel Paradox: If you break a large job into several parts, and then try to work them all at once—then each process may lose some competence, from lacking access to resources it needs.

There is a popular belief that the brain gets much of its power and speed because it can do things many things in parallel. Indeed, it is clear that many of sensory, motor, and other systems do many things simultaneously. However, it also seems clear that as we tackle more complex and difficult goals, we increasingly need to divide those problems into subgoals—and then focus on these sequentially. This means that our higher, reflective levels of thought will tend to operate more serially.

This is less of a problem for a company, which can often divide a problem into parts, which it then can pass down to separate subordinates, who can deal with them all simultaneously. However, that leads to a different kind of cost:

The Pinnacle Paradox: As an organization grows more complex, its chief executive will understand it less, and will need to increasingly place more trust in decisions made by subordinates. (See §§Parallel Paradox.)

However, many human communities are less hierarchical than the companies that we just described, and make their decisions by using more cooperation, consensus, and compromise. [205] There is usually some ‘leadership,’ but in a working democracy, those leaders are somehow given authority by the membership to help, when needed, to assist in making decisions and settling arguments. Such negotiations can be more versatile than ‘majority rule,’ which gives to each participant a spurious sense of ‘making a difference’—whereas that feeling ignores the fact that almost all differences get cancelled out. This raises questions about the extent to which our human ‘sub-personalities’ cooperate to help us accomplish larger jobs— but we don’t know enough to say much about this.

Central and Peripheral Controls

Every higher animal has evolved many resources that can interrupt its ‘higher level’ processes, in reaction to certain states of affairs. These conditions include such signs of possible dangers as rapid motions and loud sounds, unexpected touches, and the sighting of insects, spiders, and snakes. We also react to such bodily signs of aches and pains, feelings of illness, and such needs as hunger and thirst. Similarly, we are subject to more pleasant kinds of interruptions, such as the sights and smells of foods to eat, and of signals of sexual interest.

Many such reactions work without interrupting most mental activities—as when your hand scratches an insect bite, or when you move into shade to avoid excessive light. A few of these mainly instinctive alarms are: Itching, impending collision, hunger, thirst, bright light, excessive heat or cold, losing one’s balance, loud noise, pain, hearing a growl or snarl, seeing a spider, insect, or snake.

We are also subject to alarms that seem to come from ‘inside the mind’—such as when we detect an unexpected pattern or opportunity, or a failure of some process to work, or a conflict between our goals and ideals. Here are a few of these mainly acquired, internal alarms, many of which could be represented by using Critics, Censors, and Suppressors: phobias, obsessions, and sense of surprise; failure of a plan or goal; grief, guilt, shame, or disgust; and conflicts among one’s goals and ideals.

While most alarms could be handled by a Critic-Selector model of mind, one also could take a less centralized view, in which the processes that we call ‘thinking’ are affected by a host of other, partly autonomous processes. For example, one could think of a city or town as an entity whose processes are influenced by the activities of sub-departments concerned with transportation, water, power, fire, police, school, planning, housing, parks, and streets—as well as legal and social services, public works, and pest control, etc., each with its own sub-administrations. Can one think of a city as having a Self? Some observers might argue that each town has a certain ‘ambience’ or ‘atmosphere,’ and certain traits and characteristics. But few would insist that a city or town has a ‘sentient’ personality.

Reader: Perhaps that’s because they don’t have your idea that a “Self,” is a network of models, each of which may help a system to answer questions about itself. But in fact, each of those departments for planning, power, parks, and streets—and each of those other agencies—have plenty of diagrams, charts and maps that represent aspects of the town they’re in.

I cannot disagree with that—except to argue that, usually, few of those maps or models are accessible to the other departments. Perhaps if all those different representations were assembled into efficient panalogies, the resulting system might indeed seem to have more of a personality.

Programmer: I like some of your theories about how minds work—except that all of your schemes seem far too complex for a system that functions reliably enough. What happens if some of its parts break down? A single error in a large computer program can cause the entire system to stop.

I suspect our human ‘thinking processes’ frequently ‘crash’—perhaps as often as several times per second. However, when this happens you rarely notice that anything’s wrong, because your systems so quickly (and imperceptibly) switch you to think in different ways, while the systems that failed are repaired or replaced. Here are a few of the kinds of failures that are likely to get somewhat more ‘attention.’