In ancient times it was believed that the newborn mind started out just like a full-grown mind, except for not yet being filled with ideas. Thus children were seen as ignorant adults, conceived with all their future aptitudes. Today, there are many different views. Some modern theories see a baby's mind as starting with a single Self whose problem is to learn to distinguish itself from the rest of the world. Others see the infant's mind as a place containing a horde of mind-fragments, mixed together in a disconnected and incoherent confusion in which each must learn to interact and cooperate with the others so that they can grow together to form a more coherent whole. Yet another image sees the child's mind as growing through a series of layerlike construction stages in which new levels of machinery are based and built upon the older ones.
How do our minds form? Is every person born containing a hidden, built-in intellect just waiting to reveal itself? Or must minds grow in little steps from emptiness? The theories of the next few sections will combine ingredients from both these conceptions. We'll start by envisioning a simple brain composed of separate proto-specialists, each concerned with some important requirement, goal, or instinct, like food, drink, shelter, comfort, or defense. But there are reasons why those systems must be merged. On one side, we need administrative agencies to resolve conflicts between the separate specialists. On the other side, each specialist must be able to exploit whatever knowledge the others gain.
For a relatively simple animal, a loose-knit league of nearly separate agencies with built-in goals might suffice for surviving in a suitable environment. But human minds don't merely learn new ways to reach old goals; we can also learn new kinds of goals. This enables us to live within a broader range of possible environments, but that versatility comes with its own dangers. If we could learn new goals without constraint, we'd soon fall prey to accidents — both in the world and inside our own minds. At the simplest levels, we have to be protected against such accidents as learning not to breathe. On higher levels, we need protection against acquiring lethal goals like learning to suppress our other goals entirely — the way that certain saints and mystics do. What sorts of built-in self-constraints could guide a mind toward goals that will not cause it to destroy itself?
No possible inheritance of built-in genes can tell us what is good for us — because, unlike all other animals, we humans make for ourselves most of the problems we face. Accordingly, each human individual must learn new goals from what we call the traditions and heritages of our peers and predecessors. Consequently our genes must build some sort of general-purpose machinery through which individuals can acquire and transmit goals and values from one generation to another. How could brain-machines transfer things like values and goals? The next few sections suggest that this is done by exploiting the kinds of personal relationships we call emotional, such as fear and affection, attachment and dependency, or hate and love.