Nobody today on any side of the political spectrum opposes the rule of law. And for good reason. At a minimum, the rule of law carries with it the idea that legal rules should be certain, so that individuals will not be left at sea as to their legal obligations. The fixity of the law, by constraining the behavior of political actors, also improves the odds that the liberty and property of any individual will not be subject to continuous political pressures and intrigue. The standard set of requirements for the rule of law has been summarized by Lon Fuller. In his 1964 classic, The Morality of Law, he lists a variety of additional constraints, including that rules be simple enough to give clear notice of what they require, be internally consistent, have only prospective application, possess relative constancy over time, and be administered by neutral officials. (1) For these purposes, it is not necessary to resolve the many fine points that arise over the proper interpretation of this doctrine. It is quite sufficient to note that each and every one of these constraints presupposes that the language we use to express our legal rules--and hence our commands to ordinary citizens and public officials alike--is intelligible enough to meet the requirements of fair notice and neutral interpretation, even if public officials frequently violate (and are known to violate) these norms. Unless that requisite level of intellectual clarity is satisfied, then the rule of law emerges stillborn from the mouths of philosophers and legal theorists. Finguistic coherence is a minimum condition for the rule of law. Yet, as will become clear, this minimalist conception of the law does not lack any substantive component. Rightly understood, the rule of law carries more meat on its bones than an appeal to legal certainty. It has to make normative commitments to particular principles--principles, I shall argue, reflected in both Roman and common law, whose fundamental similarities are far more important than their refined differences. (2)
Both the narrow and broad conceptions of the rule of law presuppose that the tools of ordinary language are powerful enough to allow judges and scholars to formulate legal rules that make implementing the rule of law possible. Unfortunately, many scholars despair that the tools of textual analysis are not strong enough to meet the persistent challenges of the linguistic skeptic. Today, many people, both on the bench and in the academy, share this all too fashionable view of ordinary language. This undermines the rule of law, fanning the general populist unease that now infects much of our public discourse. At a theoretical level, it is common for linguistic skeptics to scoff at language as the fundamental unit of law. For example, Mark Tushnet, in his caustic review of my book Design for Liberty: Private Property, Public Administration, and the Rule of Law, celebrates the proposition that terms like "property" and "nuisance" "have no determinate content, which means that the judges must actually be relying on something else to resolve the dispute," (3) without ever letting us know what that "something else" is. This type of relativism easily extends to other terms, most notably "liberty" and "coercion," which have similarly been attacked as otiose, most famously by Robert L. Hale. In his highly influential essay, Coercion and Distribution in a Supposedly Non-Coercive State, (4) he finds that any refusal to deal should be regarded as coercive both in competitive and monopoly markets.
The danger of that position on the meaning of coercion and harm is starkly illustrated by the aggressive modern application of the idea that there is nothing wrong with limiting ordinary businessmen and women, under the antidiscrimination laws, to the choice between serving same-sex couples and losing their businesses. One notable example is the recent case of Craig v. Masterpiece Cakeshop, (5) decided by the Colorado Court of Appeals, in which the defendant cake artist, Jack Phillips, refused to prepare a wedding cake for a gay couple. One of the Commissioners, Diann Rice, said: "I would also like to reiterate what we said in... the last meeting [concerning Jack Phillips], Freedom of religion and religion has been used to justify all kinds of discrimination throughout history, whether it be slavery, whether it be the Holocaust... I mean, we can list hundreds of situations where freedom of religion has been used to justify discrimination. And to me it is one of the most despicable pieces of rhetoric that people can use... their religion to hurt others." (6)
No one who understands the meaning of the terms "coercion" or "harm" could make this statement. Nonetheless, in good Orwellian fashion, this remark was made when the Commission denied Phillips's request to suspend a commission order that required Phillips and his staff to undergo a reeducation program to make them aware of the dangers of discrimination in dealing with same-sex couples. (7) The ability to attribute coercive behavior to the victims of coercion is one dire consequence of this massive breakdown in the English language.
The dangerous point here is that the excesses of the Colorado Commission follow in linear fashion from the inability to grasp the distinction between the mass slaughter of people in gas chambers, on the one hand, and refusing to serve them in a competitive market, on the other. The phrase "using religion to hurt others" consciously elides that difference and never once asks the question of what harm the state does to individuals whom it forces by fines, injunctions, or imprisonment to surrender their religious beliefs in order to remain in business. Note the relative sacrifices. In a competitive market, dozens of other cake shops can, and will, serve this couple. But the proprietor who is forced to either go out of business or suffer reeducation has no such luxury in responding to government commands. The common law entitlements were set as they were because of the differential impact of the two sets of inconveniences, and that principle on refusals to deal is both timeless and universal, for a world with multiple alternatives is always less coercive than a world with only one. The lesson to be learnt is that whenever there is no sense of entitlements, it is easy to say that those who refuse to deal with others--say on grounds of religious conscience--are in fact engaging in coercion similar in kind, if less coercive, to the Nazi extermination of the Jews--which was of course preceded by their banishment from the common occupations of everyday life.
A second misguided way to justify the state coercion is to argue that there really is no coercion at all, for so long as people have been left with a choice, they cannot be said to have suffered from any form of legal duress. (8) Making wedding cakes or going out of business are the supposed choices left to Jack Phillips. But as Justice Holmes said long ago: "It always is for the interest of a party under duress to choose the lesser of two evils. But the fact that a choice was made according to interest does not exclude duress. It is the characteristic of duress properly so called." (9) Again, no account of duress is intelligible without a robust account of the antecedent property rights. There is no duress if your choice is between my watch and your money. There is duress par excellence when your choice is confined to surrendering either your watch or your money. The analytics are what make the argument go. The self-conscious decision to put key terms in quotation marks is sure evidence that the sound understandings of ordinary language have been rejected or ignored, opening the path to near-totalitarian excesses. Reeducation is a dangerous word worthy of Pol Pot's Cambodia and Mao Tse-tung's China.
It is linguistic shifts toward skepticism similar to, if less calamitous than, the reasoning against religious cakemakers in Colorado that in large measure have fueled the expansion of government power from the classical liberal tradition to a modern progressive one. This change has proceeded in two directions at the same time. First, the constitutional limitations on the legislative power of the federal government have eroded with an expansive reading of the Commerce Clause and a diminution in the protections afforded to private property and economic liberty. Second, the opposite result has occurred with the unjustified expansion of judicial power that expands the constitutional limitations on the state's ability to regulate abortion and impose punishments on various criminal offenses. The case for judicial control is stronger on the scope of federal power and the protection of property entitlements than it is in setting the definitions and punishments of various criminal offenses. In the first two cases it is possible to set clear boundaries on government conduct, which is far harder in the last two cases given that setting punishment involves an uneasy amalgam of concerns with deterrence, retribution, rehabilitation, and incapacitation.
These major political interventions were driven by a view of language that marks a sharp departure from the Framers' confidence that the English language was clear enough to organize the fundamental institutions of government. They knew what it meant to divide government powers into the legislative, executive, and judicial branches, as is done in Articles I, II, and III of the original Constitution. They did not unduly trouble themselves with the inevitable disputes that arise at the margins of these categories, much less use borderline cases (of which there are always many) to undo the basic contours of their system, which features first separation of powers and then the checks and balances among the various branches. All too often, however, the writers in the progressive tradition (who first ushered in and then defended the Court's New...