Suppose that, on some morning, I intend to get to the university. This is my immediate purpose. However, it is subordinate to some higher goal, for instance to teach some seminar. The seminar, in turn, is not an ultimate goal, either; instead, its purpose is to transmit the true linguistic theory to the next generation. This, in turn – and skipping one or another intermediate level –, is an essential prerequisite for the highest goal in my life, which is to attain eternal bliss. This shows that acts and actions have a position in a hierarchy where each but the highest goal is subordinate to a higher one.

This hierarchy continues downwards, too. Just as I am free to choose among various activities that lead me to eternal bliss, so now I can choose among various means that get me to the university. I choose going by bicycle. As I am riding it, I have to pedal. To do this, I press my front foot down while releasing my back foot. In order to press the front foot down, I have to contract certain muscles. In order to contract them, I have to send them a certain neural signal. At the bottom, this ends in certain elementary electrochemical processes.

This gives us a hierarchy in which a given action is a means for a superordinate action which is its purpose, and simultaneously functions as the goal for a subordinate action or process which serves it as a means. Such a means-end hierarchy is a teleonomic hierarchy. A section of it comprising levels n-1 to n+1 takes the form shown in the Diagram.

Teleonomic hierarchy
purpose act/process of level n+1
means act/process of level n purpose
act/process of level n-1 means

At the bottom of the hierarchy there are processes developing automatically and not directly controllable. At its top, there are goals which are set freely and consciously. Between the top and the bottom, there is a transition of several levels whose nature is shown in the following table.

Levels of a teleonomic hierarchy
   features
level     ╲
complexity control automation
higher higher higher lower
lower lower lower higher

At any given level of a complex action, choice of an act obeys two conditions:

Illustrating with the above example:

Thus, typically there are at each level isofunctional strategies which fulfill the given function in similar ways and, in principle, equally well.

In addition, a given strategy is often polyvalent (or multifunctional) because it can also be used for other purposes. For instance, both bicycle riding and walking, but not taking the tramway are suitable for physical training. Therefore, if I want to achieve more than one goal at a time, choice of a suitable means is also determined by which of them serves more than one of my goals at the same time.

It is not the case that a certain degree of control was assigned, once for all, to a certain level of a teleonomic hierarchy or to a certain action or process. Instead, within certain limits, the degree of control with which a given action or process is executed is variable. This does not include the very top and the very bottom of the hierarchy. Thus, on the one hand, there is no known automatism which would directly attain the goal of eternal bliss. And on the other hand, no method is known by which a person could control directly – i.e., without the intercalation of aids like medicine – his digestion. However, at intermediate levels, control is shiftable. For instance, in biking, I generally do not control consciously my pedaling. However, if am instructed by my biking coach to change my pedaling habits, I can control it. Again, this possibility is narrowly limited, as will be seen shortly.