The only way an autopoietic system can fail is if it stops reproducing itself. Instrumental purposes are constructed by another observing system. What we might consider success for a system, such the education system transforming all human beings into persons, would lead to the education system’s dissolution or self-annihilation.
For instance, if we think education’s purpose is to transform human beings into persons (person, as distinguished from human beings, can access all function systems), then education’s complete success would cause the end of the education system because it would have nothing left to do. Thus, if a certain, predictable percentage of students do not graduate or gain the competence to access the economy, politics, etc., that doesn’t mean the education system has failed.
Dropping out of school or failing an exam or a class must always be possible. Student success and failure is a two-side form. Both sides of the form must be accessible; that is to say, the boundary between the two sides of the form must be crossable.
It is not morally bad to lack certain skills or knowledge that schools can teach. Failing test or even failing a grade is not a moral issue. Incompetence or lack of understanding of some part of school curriculum is necessary. In fact, education must continually find new ignorance or incompetence, which it then tries to replace with knowledge or skills. Ignorance is produced in the same moment that competence is established. The two sides of the code are always applied at the same time to the same situation. The new learning produced by education has to open up new ignorance or incompetence. This is where selection comes into play. Assessments, including exams, are meant to distinguish incompetence and competence (or mastery/non-mastery), or to reproduce the distinction. The education system must keep inventing new kinds or level of ignorance, which it can then try to remove. Otherwise, the education system has nothing to do.
New areas of study or research fields also construct new ignorance. There is always some new ignorance to construct (e.g., sexism, racism, xenophobia, homophobia, or lack of understanding of climate change) and the education system, which is not limited to schools, can then get to work trying to eliminate these new forms of ignorance. Ignorance is socially constructed; it’s not a thing-in-itself. It’s a difference–the other side of knowledge.
The same applies to the health system. It must continually construct new illnesses that it can treat. A health system that cannot create new diagnoses will stagnate. Thus, conditions that in the past were just considered part of human life and not even given a name are identified and given names (e.g., Alzheimer’s disease, major depression, obsessive-compulsive disorder, PMS, erectile dysfunction, restless leg syndrome, insomnia, low testosterone); they are classified as treatable illnesses in order to give the health system something to work on.
The purpose of the economy may be said to be distribution of scarce resources or maybe to solve the problem of scarcity so that everyone has “enough” of what they need to live comfortably. Be these instrumental goals are created by some other system, such as politics or some kind of moral system or organization; they aren’t the goals of the economy because the economy has no instrumental goals. It just circulates payments. If it loses the capacity to circulate payment, then it fails.
Failure is an observation by another system. For example, . Attempts to solve scarcity problems must create new scarcity problems. Regarding markets, Steffen Roth (2016) writes,
The most constructive attitude to market failures we can cultivate is that we accept (our) market failures as necessary conditions for the existence of markets. Even if we started from the instrumental assumption that markets have a purpose and that the purpose of economic markets is to solve problems of scarcity, then we may realize that the idea of a perfectly instrumental, ideal market ultimately implies the idea of the (self-)annihilation of that market. In fact, markets can only be sustained if attempts to solve scarcity problems ultimately create new scarcity problems. An ideal market that does not eventually fail to fulfill whatever its supposed purpose may be would therefore cease existence. In that sense, we may either embrace market failure or engage in the destruction of markets. A positive attitude to market failure therefore is the only constructive attitude to markets.
Similarly, educational failures are necessary for the continued existence of the education system, and scientific failures are necessary for the continued existence of science.
What we consider system failures are actually disappointed expectations. If millions of people die of cancer, that is not a failure of the health system, or if millions of people starve to death while thousands of new billionaires arise, that is not a failure of the economic system. If 50% of students do not graduate from high school, this would be a major disappointed expectation but not a system failure. It could be called a system failure by some system, such as politics or the economy.
These outside systems can be interaction systems, organizations, or function systems. Interaction systems arise between people in direct, fleeting communication, whether face-to-face or technologically mediated (despite what Luhmann says about mediated communication). This is where empathy may come into being. Organizations can also form to alleviate the suffering that results from the economy, the education system, the health system, or some other system. The political system or the religious system can also come into play.