Politics, Thermostats, and Existentialism

Control theory has been introduced in previous posts, in the context of explaining or describing the nature of the complexity of our political systems, and specifically whether an existential orientation towards politics is justified or not.

If causes and effects can be reliably identified and consistently traced through our political systems, then the existential assumption that our political systems are irreducibly complex is invalid, and therefore an existential orientation towards politics is unnecessary. However, if our basic assumptions about how our political systems work can be shown to be fatally flawed, particularly pertaining to our understanding of what our political systems control or do not control, then conventional conceptions of citizenship based on these assumptions of coherence and control will be fundamentally inadequate—or, worse, actively contributing to the apathy and pessimism which characterize so much of our contemporary politics. As such, because Existentialism is explicitly premised on taking meaningful action in an implacable world that repels ultimate meaning, an existential account of citizenship is the only orientation towards politics that can not only embrace this incoherence but still motivate meaningful and enduring political engagement.

This installment will thus demonstrate one such fundamental challenge to our unquestioned assumptions of the ultimate coherence of our political systems from the application of control theory, and why an existential approach to citizenship is the most appropriate response.

What the thermostat knows

As described in more detail in previous installments, a control system is any system whose components act together in such a way as to stabilize against disturbances some quantity or property in the ambit of the system. Notably, there is only one basic way that control can be exercised. To the degree the components of a system perform the functions described by control theory, and relate to each other in the order described by control theory, that system—regardless of its physical composition or complexity—will be exerting control over the quantity or property in its focus.

The prototypical example of a control system is the common household thermostat:


Believe it or not, even at this rudimentary level of explanation a proper understanding of how control systems must function to exert proper control already reveals a fundamental misunderstanding of political control which is evident at all levels of society, from regular people, to politicians, to academics: The one basic aspect of control systems that most people confuse—and which results in fundamental errors in the attribution of control, which is such a critical aspect of political systems—is in their identification of what is the focus or target of control.

As demonstrated above, for the thermostat to function as such there must a component which senses the ambient temperature of the room as the input, converting that into a feedback signal. This feedback is then compared against a reference signal (i.e., the setting of the thermostat), returning the magnitude of the difference between the present room temperature and the desired room temperature as an error signal. This difference is then translated by the output function into a change in the state of the furnace (off or on, or no change if the error signal is within the parameters). The action of the furnace then contributes to the temperature of the room.

However, the temperature of the room may in turn also be subject to other influences in the external environment (open doors, summer or winter weather outside the house, a crackling fire in the room, etc.). As such, the temperature of the room—as the combination of the results of the output of the system and any environmental perturbations—is then sensed by the feedback function, compared against the reference, and on and on. Again, only by these specific components performing these specific functions in this specific order does the thermostat fulfill its intended purpose of controlling the temperature of the room.[1]

This likely seems a basic, obvious, and superfluous example, with little relevance to the question of control in and by political systems. For as trivial as this example might seem, though, this demonstration of the sequence of functions and relations which are necessary for control suddenly becomes very significant when this basic understanding of control is applied to politics.

Fundamental misconceptions of control

Keep in mind that if the components of a system do not perform these necessary functions in the proper relation to the other components, even the simple thermostat will not be able to control the temperature of the room as intended. If the comparator compares the reference signal to the output signal, or the feedback function processes the error signal instead of the input quantity, then this chain is broken and the room temperature is not stabilized or controlled.

However, when asked to describe how a control system operates, even one as basic as a thermostat, most people will say that what is being controlled is the output of the system.[2] This makes superficial sense, as this output is the ‘product’ of the system, and therefore what seems to be the purpose of the system, but this is a fundamental mistake. Notably, this is the same fundamental mistake that many people also make in their attribution of control to political systems.


As demonstrated in the basic thermostatic model above, control systems do not stabilize their outputs; rather control is produced by focus on inputs (i.e., true control systems control their inputs, not their outputs).[3] The outputs are necessarily related to the inputs, otherwise there would be no cause-effect relationship to be manipulated for the purposes of control, but, contrary to these appearances, true control systems do not actually control their outputs. This distinction is exceedingly important in the attribution or assumption of control in and by political systems.

Consider a thermostat that did control its outputs in the way most people assume. Such a configuration of control would necessarily look like this:


Even this rudimentary example demonstrates the significant issues which arise from this common misconception of control. For one, as can be seen, this would mean that the action of the thermostat is indifferent to the temperature of the room, and is concerned instead only with the output of the system (i.e., the level of heat produced by the furnace). Another consequence of this misconception of control is that the reference signal would now also have to pertain to the amount of heat produced by the furnace, and not to the temperature of the room. The end result is that such a system would control the level of heat coming out of the furnace, and not the temperature of the room, which is the presumed purpose of a thermostat.

Political systems: Out of control?

Now consider control conceived in this way—as the control of outputs and not inputs—applied to a political system, such as a policymaking agency, for example. Assuming that all the necessary components have been identified, and that they perform the requisite functions in the proper sequences, this configuration of control would look like this:


From this configuration of control based on this common assumption that political systems control their outputs, there are a couple of conclusions which have unnerving implications for our understanding of politics, and therefore for our motivations for meaningful political engagement:

  1. If political systems actually do control their outputs in this way, they would be closed systems fundamentally disconnected from their political environments. In this case, these political systems would be exerting control, but only of themselves and not in response to ongoing changes in some aspect of society at large that we assume are the purpose of our political systems.
  2. If this common conception of control is instead a misconception of control, and our political systems are in fact controlling their inputs in the proper way, then this means that most people fundamentally misunderstand how our political systems actually work. In this case, people are assuming that our political systems are controlling something that they are not. The subsequent problem is that this means we do not actually know what our political systems are controlling—they may well be controlling some aspects of social life, but the problem is that we have not yet identified what those aspects of society are.

The upshot from this rudimentary demonstration is that either we do not understand how our political systems actually function, or our political systems do not function as we assume they do, specifically in terms of what they do or do not control, which is the primary justification for our political institutions and agencies.

This fundamental misconception of how our political systems actually work is one reason why the first and second principles of the Existential Citizenship Project are so important, because otherwise the result is the pervasive and uncritical acceptance of more or less mythological accounts of how our political systems operate. To the degree that our explanations of political systems do not explain how they actually operate is also the degree to which prescriptions for action based on these flawed explanations are liable to mislead and disappoint, leading to widespread apathy or nihilism. Because we either do not understand how our political systems work, or because they fundamentally defy explanation, what is needed are justifications for political actions that do not rely on the coherence of the system or on our comprehension of it. Again, these fatal flaws in our basic understanding are why an existential approach towards politics is the most apt orientation for political engagement.

Existentialism and political control

That said, there is also still the possibility that our political systems do in fact control their inputs in the proper way, as described by control theory. If this is the case, then we might be able to comprehend the system after all. The basic diagram of political control properly conceived in this case would look like this:


Again, as a proper control system, this means that an input quantity is detected by some feedback function within the policymaking entity, which is then compared in some way against a preferred ‘level’ of policy, generating an error signal of the difference that initiates an appropriate output. In the context of a policymaking entity, this output would presumably be a level or quantity of policy.

However, even if this is how our political systems actually work, there is still a substantial problem revealed by this control theory approach that does not yet appear to have a satisfactory resolution: What are the inputs into such a system (i.e., what is the actual controlled quantity)?

Again, this controlled quantity would not be the ‘level’ of policy, as this is the output of the system itself, which would result in the self-referential ‘closed thermostat’ model diagrammed above. So this input must be something else. Given how we conceive of our political institutions, these inputs would ideally be some effect in society at large as the focus or target of policy. There are any number of possibilities for these controlled quantities, depending on the political system being analyzed (Public preferences? Political mandates? Bureaucratic imperatives? Etc.), but the identification of these inputs is ultimately an empirical matter.

To my knowledge, though, there has been very little work done on identifying the inputs into political systems as a focus of control as suggested by control theory.[4] I have yet to see the inputs of specific policymaking systems—when they are discussed at all—treated as the proper focus of control, much less the inputs for political systems as a whole. This absence of attention in the literature to inputs as controlled quantities strongly suggests, again, that we do not as yet know what our political systems actually control, which means that even with a proper conception of control we do not yet adequately understand how our political systems actually work. This is in addition to the fact that the general conception of political systems as controlling their outputs is still a fundamental mistake, and therefore that most people—including most politicians and academic scholars—still fundamentally misunderstand how our political systems as systems of control actually work.

Given these fundamental flaws in our basic understanding of how our political systems work, even with a proper conception of control, our traditional justifications for political engagement based on these assumptions of coherence and control are at best inadequate and at worst actively contributing to the apathy and pessimism which characterize so much of our contemporary politics. What is needed instead is a conception of citizenship capable of acknowledging this fundamental incoherence of our politics and yet is still able to motivate meaningful and enduring political engagement. The only philosophical stance explicitly premised on taking meaningful action in a world that defies our understanding is Existentialism. This is why an existential account of citizenship is the only viable orientation towards politics in the face of all this fundamental uncertainty.

[1] Another simple example of the importance of the focus of control systems on their inputs and not their outputs is to think of a driver steering a car as a control system. In this case, the inputs are the events outside the windshield as the focus of the attention of the driver, the reference signal is provided by the lane markers, and the desire of the driver to keep the car in its proper lane and to avoid obstacles, and the outputs are the motions of the hands of the driver on the steering wheel which affect the position of the car. If the driver were to be controlling their outputs, as is the common assumption about control systems, the driver would be focusing on their hands on the steering wheel, and on keeping those motions within certain parameters, which is obviously a disastrous way to drive a car. Instead, by focusing on the proper inputs, the driver is able to respond effectively to even unexpected events using their outputs to adjust the orientation of the car according to these inputs.

[2] This evidence is admittedly anecdotal from my own experiences over years of asking people seemingly inane questions like “So what do you think a thermostat controls?”, but this misunderstanding of basic control principles has been my repeated experience. By all means go ahead and try this out yourself, and let me know what you find out.

[3] In well-functioning control systems, the outputs may often behave in a consistent manner correlated with the inputs that might appear to be control, but this is because the outputs are a systemic response to the inputs—not because the outputs themselves are being controlled (i.e., as the inputs change, so also will change the outputs). On the other hand, because the system is responding to both its own outputs and perturbations in the environment, the outputs can sometimes vary widely from one moment to the next, depending on the magnitude of the error signal that is generated. However, in a proper control system, these wild swings in output will be contrasted with a relatively steady input signal—again, because what is being controlled are the inputs and not the outputs, this seeming lack of correlation of inputs with outputs is in fact a sign of true control. What is more important in assessing the presence or absence of control is the relation between the inputs and the reference signal.

[4] At least in the domain of public policy, which is the area I have looked for control in the most. In other areas of research in political science and theory it may be that political control is properly conceived as control of inputs, and is therefore studied appropriately, but I have not yet encountered evidence of this.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s