Autopoietic systems reduce environmental complexity—narrow the range of potential stimuli—in order to solve problems. And is doing so, systems build up their own internal complexity—or, one might say, their resources for survival. As Luhmann puts it,
“[Evolutionary] advances reduce complexity in order to organize greater complexity on the basis of restriction. Thus a road network reduces the possibilities for movement to enable easier and faster movement and hence increase options for movement concretely available.”(2012, p. 306 )
Many cities are, of course, built on a grid design for this reason. The grid restricts automobile and pedestrian movement while enhancing the efficiency of movement. Scoppa, Bawazir, & Alawadi (2018) analyzed the walkability of ten “superblocks” in Abu Dhabi and found that a complex network of the narrow alleyways, or sikkak, “make remarkable contributions to the walkability of superblocks. In particular, sikkak turn many of the fragmented networks which characterize Abu Dhabi’s superblocks into highly efficient layouts which resemble dense orthogonal grids” (p. 367). Efficiency is often desirable; however, grids are often cages. Furthermore, the need to maintain or increase efficiency places a burden on traffic engineers, social engineers, and all kinds of grid-makers.
A grid is a sort of complexity-reduction machine and, like any machine, it must be maintained. These machines also need to be continually improved as weaknesses, or gaps in the grid, become apparent. The problem is that grids designed to increase social order (e.g., traffic grids, the rows and columns of a classroom) face resistance from operationally closed autopoietic systems. There has never been a panopticon or disciplinary machine built that can control operationally closed social or psychic systems. A physical structure such as high walls, narrow hallways, and prison cells do certainly discipline bodies, and they can also block communication; however, these structures cannot discipline minds.
Discipline works best when those subjected to it have options—in other words, when the desired response is not coerced: An effective urban grid will not have sidewalks that prevent a pedestrian from turning around to retrieve the keys she accidently left at a diner. If we envision a continuum with the absence of discipline on one end and coercion on the other end, disciplinary mechanisms work best somewhere in the middle.
Grids and other disciplinary mechanisms are designed to reduce contingency—i.e., unlimited possibilities or too many options to select from. The one subject to discipline should have a reasonable number of options, and the one directing the discipline should also have multiple options. But the closer the disciplinary mechanism moves to the coercion side of the continuum, the fewer options the disciplinary apparatus has. At some point, for example, the prison system just locks a prisoner up twenty-four hours a day and feeds him through a slot in the cell door. The burden of discipline—i.e., the task of reducing contingency—then falls wholly on the shoulders of the disciplinary apparatus; the one subject to discipline is, paradoxically, released from any obligation to share in his own discipline. As Luhmann puts it,
the reduction of complexity is not distributed but is transferred to the person using coercion(Trust and Power, 1979: 95)
To cope with disciplinary challenges, entire new scientific fields have emerged. For instance, the invention of prison led to the development of a new science—prison science, which Henderson (1911) defined as the
systematized knowledge which is required as the basis for the treatment of offenders in institutions of correction or in connection with their administration.
A recent database search for “prison science” returned 164 peer-reviewed articles. A search for “prison management” returned 2,151 peer-reviewed articles. Similarly, classroom management is an entire subfield of the “science of education.” A search for “classroom management” returned 23,227 peer-reviewed articles. The examples are potentially endless, but the point is that efforts to reduce complexity in one place generate new complexity somewhere else, sometimes in the science system.