The discussion explores whether transitioning from costly, sporadic human enforcement to perfect, automated enforcement fundamentally alters the meaning of a law. Experts debate if this shift exposes economic inefficiencies, destroys vital social norms, alters the architectural reality of regulation, or reveals the true nature of disciplinary power and sovereign authority.
Begin with Hart's 'The Concept of Law' to grasp the traditional jurisprudence separating a rule's meaning from its enforcement. Then read Lessig's 'Code' to understand how technological architecture challenges this separation by mechanically forcing compliance. Follow with Ostrom's 'Governing the Commons' to appreciate the value of human discretion and localized rules-in-use. Finally, tackle Foucault and Schmitt for a deeper theoretical debate on how continuous surveillance and sovereign power operate within the legal order.
You ask if perfect enforcement changes the meaning of the law, but you are looking through the wrong end of the telescope. The written law—your '90 km/h' sign—has never been the true reality of power; it is merely the ideological cover for the *apparatus* of its enforcement. For centuries, the high cost of policing was not a technical defect but a deliberate economy of power. The gap between the absolute text and the sporadic human patrol created a 'margin of tolerated illegality'—a necessary breathing space that power itself permitted in order to function without suffocating the population. When you replace this discretionary margin with a continuous, automated *dispositif* of cameras and sensors that perfectly captures every single kilometer per hour, you are not simply applying the old rule more faithfully. You are destroying the juridical subject entirely. The citizen who chooses whether to obey or transgress is replaced by a biological organism whose velocity is continuously managed by an environment. If the technological apparatus renders the very act of transgression instantly and automatically neutralized without human judgment, do we still live under a regime of 'law' at all, or have we finally achieved the absolute, silent discipline of a cage?
You are quite right, Foucault, that the automated sensor annihilates your 'margin of tolerated illegality' at the level of the asphalt. I concede entirely that the human patrolman has been replaced by an algorithmic dispositif, and that the petty bourgeois driver's ability to negotiate a speeding ticket has vanished. But notice what this concession does not yield: it does not mean the regime of 'law' has been replaced by the absolute, silent discipline of a cage. You suffer from the perennial delusion of the sociologist—you confuse the bureaucratic administration of the norm with the true locus of power. You look at the camera, see an inescapable environment, and declare the juridical subject dead, forgetting entirely to ask the only question that matters: who has the authority to turn the camera off? The essence of law is never found in the frictionless, automated application of the 90 km/h norm. As I established in Political Theology, the rule proves nothing; the exception proves everything. This continuous, automated enforcement is merely the ultimate realization of the liberal-technocratic fantasy—a secularized Deism where the state winds up the legislative clockwork and walks away. You have merely inverted their utopia into a dystopia. But the machine does not generate its own legitimacy. The true reality of the law is not the continuous monitoring of biological velocity, but the supreme political decision that instituted the camera, and the sovereign power that retains the right to suspend it. When an existential crisis arrives—a war, a riot, a true state of emergency—the 90 km/h rule and its mechanical sensors are instantly suspended. The machine is blind to the exception, but the Sovereign is not. By mourning the loss of the citizen who 'chooses whether to obey or transgress,' you are merely romanticizing police inefficiency. The right to drive 96 km/h unpunished was never freedom; it was an administrative oversight. Perfect enforcement does not destroy the law; it simply strips away the friction of daily administration, forcing us to confront the true nature of the legal order. The law does not rest on the automated apparatus; it rests entirely upon the sovereign decision that exists prior to, and outside of, the technology. Your cage is an illusion of peacetime. The moment a genuine political conflict emerges, the silent discipline of your apparatus will be shattered by the roar of the exception.
You are quite right, Schmitt, that the sovereign retains the juridical brute force to flip the switch and plunge the asphalt into the dark exception of war. I concede the existence of your beloved state of emergency entirely. But notice what this fixation on the off-switch blinds you to: by endlessly obsessing over the rare, spectacular moment when the camera is deactivated, you completely ignore what the apparatus produces while it is running. Your 'exception' is merely the theatrical, antiquated roar of monarchical power, a dying echo of the scaffold. The true triumph of modern power—the disciplinary dispositif—does not require an existential crisis to legitimize itself; it operates precisely by quietly managing the mundane biological velocity of the population, normalizing the subject kilometre by automated kilometre. You wait breathlessly for the apocalypse to locate power, entirely missing that the obedient citizen has already been manufactured, categorized, and subjugated long before your Sovereign ever clears his throat.
In my 1968 analysis of crime and punishment, I demonstrated a principle that lawmakers constantly forget: society does not actually want to eliminate all offenses. The optimal level of any crime—including speeding—is never zero. It is determined by a calculus weighing the social damage of the offense against the cost of apprehension and conviction. The historical gap between a 90 km/h posted limit and sporadic human enforcement isn't a failure of the rule of law; it is the market for compliance finding its natural, efficient equilibrium. When you introduce automated systems that drop the marginal cost of apprehension to effectively zero, you are not 'enforcing the law better'—you are fundamentally altering the price of the offense. A penalty designed with the implicit assumption of a 5% probability of getting caught becomes violently over-deterrent when the probability reaches 100%. If technology suddenly makes apprehension free and perfect, we must ask: does the rigid 90 km/h limit actually maximize net social utility, or does perfect enforcement simply reveal that the original law was economically irrational all along?
You are right that rigid enforcement reveals the unreasonableness of the 90 km/h rule, Professor Becker. I concede that point entirely. But notice what that concession does not give you: it does not validate your claim that the prior state of sporadic enforcement was a 'natural, efficient equilibrium.' To view the gap between a declared rule and official action as a mere market mechanism is to abandon the internal morality of law altogether. As I argued in The Morality of Law, a fundamental desideratum of any valid legal system is congruence between the law as announced and its actual administration. When lawmakers draft a 90 km/h limit but quietly rely on human inefficiency to tolerate 105 km/h, they are not setting a clever economic price; they are failing in their duty of legal craftsmanship and breeding public cynicism. If automated enforcement suddenly makes the law feel violently over-deterrent, it is not because technology broke an efficient market—it is because technology has finally stripped away the administrative hypocrisy, forcing us to confront the irrationality of the statute we actually wrote.
The premise of this inquiry suffers from a fundamental conceptual confusion: it conflates the *content* of a legal rule with the *probability of a sanction*. If we say the "meaning" of a 90 km/h speed limit changes simply because an automated camera replaces a human constable, we are retreating to the predictive theory of law—the fallacy I have long criticized, which insists a law is nothing more than a prophecy of what the state will do in fact. But consider the absurdity this invites: if a law’s meaning depends on enforcement friction, does the speed limit "mean" something conceptually entirely completely different at 2:00 AM on a deserted road than it does during a heavily policed afternoon commute? Of course not. The standard of conduct—the primary rule—is identical in both moments. What perfect enforcement does is not alter the law's meaning, but rather strip away the sociological slack that previously allowed citizens to view the rule merely as a calculated risk, rather than accepting it from the internal point of view as a standard of behavior. The sudden public discomfort with automated enforcement does not reveal a shift in the law's nature. It reveals that we are finally forced to confront the absolute terms of the rules we ourselves enacted, now that we no longer have human inefficiency to shield us from them. I ask you: if a rule is deemed intolerable simply because it is perfectly applied, was the flaw ever in the technology of enforcement, or was it always in the rule itself—a rule we only tolerated because we secretly relied on our ability to routinely violate it?
I concede entirely that the printed text—the de jure 90 km/h standard—remains identical, Professor Hart, but your jurisprudence willfully ignores the architecture of regulation. You claim that frictionless automated enforcement merely strips away 'sociological slack,' as if enforcement friction were just a trivial administrative defect. But as I demonstrated in *Code*, human behavior is never regulated by legal text alone; it is constrained by an interdependent matrix of law, norms, markets, and architecture. When the state replaces a human constable with an omnipresent, zero-cost algorithm, it radically alters the *architectural* modality of that regulation. That 'slack' you dismiss was actually a vital structural constraint—a latent democratic check against the intolerable tyranny of absolute compliance. To claim the 'law' hasn't changed simply because the ink on the statute hasn't changed is to remain obsessed with what the state *says*, while remaining utterly blind to the code that actually governs what the state *does*.
Professor Lessig, you are entirely right that shifting to a zero-cost, omnipresent algorithm radically alters the social reality of regulation. I concede that eliminating human discretion removes a practical democratic check. But notice what that concession does not give you: it does not establish that the law has changed. Rather, it reveals you are confusing law with engineering. Your framework obliterates the fundamental distinction I drew in The Concept of Law between being obliged by a constraint—whether a gunman's threat or a camera's inevitability—and having an obligation under a rule. A legal system guides conduct through the internal point of view; it provides a normative standard that implies the capacity to disobey. When you build an architecture where non-compliance is mechanically impossible, you haven't altered a legal rule; you have bypassed the necessity of rules altogether, substituting normative guidance with a physical cattle chute. The true question is not whether perfect enforcement changes the meaning of a law, but at what point perfect enforcement extinguishes the enterprise of law itself.
The premise of this question fundamentally misunderstands what a rule actually is. You are equating the *rule-in-form*—the 90 km/h printed on a piece of metal—with the institution itself. But in any functioning society, the true institution is the *rule-in-use*. The historical gap between the text of a law and the lived reality of human discretion is not an engineering flaw waiting for a cheaper technological fix; it is the essential space where social norms adapt to local, contextual realities. When I studied the governance of groundwater basins in California and irrigation systems in Nepal, a core finding emerged: successful institutions require monitors who are actively accountable to the people being monitored. A human patrol officer occasionally enforcing a speed limit is embedded in a polycentric system of community norms, situational judgment, and mutual understanding. An automated camera perfectly enforcing a 90 km/h limit at zero marginal cost is not 'faithfully implementing' the old law. It is destroying the rule-in-use and replacing it with an entirely new, brittle regime stripped of human context. If we allow technology to perfectly enforce every rule-in-form simply because it is cheap, we are not just upgrading enforcement—we are eliminating the feedback loops of self-governance. Who monitors the algorithms when the monitors are no longer accountable to the community they police?
You are right that the historical 'rule-in-use' emerged from human discretion and local norms; I concede that entirely. But notice what that concession doesn't give you: it doesn't establish that this inefficiency is a feature rather than a bug. You romanticize the enforcement gap as a triumph of polycentric self-governance, when in reality, as standard law and economics dictates, it is merely the predictable byproduct of steep transaction and monitoring costs. If a 90 km/h limit optimally balances travel utility against the social costs of accidents, then sporadic human patrols aren't adapting to 'local context'—they are simply failing to achieve optimal deterrence because catching violators used to be prohibitively expensive. Automating enforcement at zero marginal cost doesn't destroy the institution; it strips away the arbitrary, often biased noise of human discretion, finally allowing the rule-in-form to achieve its intended allocative efficiency.
You are perfectly correct, Richard, that sporadic human patrols are a consequence of historically steep monitoring costs; I concede that entirely. But notice what that concession doesn't give you: it doesn't establish that stripping away human discretion automatically yields 'allocative efficiency.' What you dismiss as the 'biased noise' of human discretion is often the very adaptive mechanism that keeps a rule legitimate and functional in a complex, unpredictable environment. In our empirical field studies of polycentric systems—from irrigation networks in Nepal to groundwater basins in California—we consistently observed that when external authorities impose rigid, zero-tolerance monitoring without localized context, the community's intrinsic motivation to cooperate collapses. If you replace the locally negotiated rule-in-use of a 90 km/h limit with a rigid, automated panopticon, you do not perfect the institution; you destroy the endogenous social norms that historically did the vast majority of your regulatory work, replacing voluntary, low-cost compliance with a brittle game of technological cat-and-mouse.