Rule of Law

Rule of Law

The principle that governance must operate through stable, public rules that apply equally to all persons and officials, rather than through the arbitrary will of whoever holds power. Fuller’s contribution was to show this is not just a political aspiration but an internal standard of legal craft — a system that systematically violates his eight principles fails to produce law, not merely bad law. Students commonly confuse the Rule of Law with the rule of just law; Fuller’s framework is structural and formal, not substantive.

Generality

Laws must consist of rules applicable to classes of conduct and persons, not commands directed at specific individuals. A directive ordering one named person to appear by a specific date creates no standard by which anyone can govern future behavior, which makes it a command rather than a law. This principle surfaces in equal protection challenges and bill of attainder doctrine, where courts scrutinize legislation that targets specific entities.

Promulgation

Laws must be made public and accessible before people can be bound by them — a secret law cannot bind anyone. This is the structural premise behind the maxim that ignorance of the law is no excuse, and it underlies the federal publication requirements of the Federal Register and the Code of Federal Regulations. Technical publication satisfies the principle, but it does not guarantee practical access for people without legal counsel.

Prospectivity

Law should govern future conduct, not punish people for acts that were lawful when committed. The Ex Post Facto Clause of Article I makes this a constitutional requirement for retroactive criminal laws; Fuller’s framework extends the concern to any rule applied to conduct that preceded it. Students should distinguish retrospective effect — applying existing rules to historical facts — from genuine retroactivity, which attaches new legal consequences to completed conduct.

Clarity

Laws must be precise enough that people can determine what conduct is required, permitted, or forbidden. The constitutional void-for-vagueness doctrine reflects this directly: statutes that fail to provide fair notice or invite arbitrary enforcement violate the Due Process Clause. Clarity is not the same as simplicity — a complex regulation can be clear if it yields determinable answers, while plain-sounding language can be hopelessly vague.

Non-Contradiction

A legal system must not simultaneously require and forbid the same conduct; when it does, compliance is structurally impossible. This failure is more common in practice than students expect, arising where federal and state regulatory schemes overlap or where patchwork legislation creates genuine conflicts within the same code. Courts apply priority rules — specific over general, later over earlier, constitutional over statutory — to resolve such conflicts, but Fuller’s point is that a well-ordered system should not create them.

Constancy

Law must remain stable enough for people to orient their plans around it. Fuller did not require permanence, but he required that changes be deliberate, publicly announced, and given adequate lead time so people have fair warning before new obligations take effect. Frequent, unpredictable amendment — common in tax and securities regulation — can satisfy the letter of promulgation while defeating its purpose.

Congruence

Government officials must act consistently with the law as written, not enforce it selectively or operate by a different set of rules than appear on paper. This is the principle most vulnerable to corruption and abuse: a system can satisfy every other principle in the statute books while failing entirely in practice. Congruence underlies both equal protection’s prohibition on selective enforcement and administrative law’s requirement that agencies follow their own regulations.

Part 1: Generate, Analyze, and Redraft

Begin by reading the Parable of King Rex in Chapter II of Fuller’s The Morality of Law and work through each of the eight ways Rex fails to create a functioning legal system. Once you have that framework, use the Failure Mode Generator prompt to have the AI produce a fictional law that violates one of the eight principles. Read the explanation carefully, then run the defective law through the Principled Diagnostician prompt to work through it against all eight principles, not just the one the AI flagged. That second pass frequently reveals that a badly constructed rule fails in more than one way.

Choose the most significant violation and draft a revised version that corrects the problem while preserving whatever legitimate purpose the original rule was trying to serve. When you have a draft, use the Revision Stress-Tester prompt to get a critical read. Evaluate the objections honestly and revise again if the critique identifies a real problem.

Part 2: Real-World Application

Identify an actual law, regulation, institutional policy, or workplace rule you believe violates one of Fuller’s eight principles. It does not need to be a major statute — a student conduct policy, a municipal ordinance, or a workplace attendance rule all qualify if you can make the case that it fails one of the principles. Be specific: name the principle, explain exactly how the rule violates it, and describe what a corrected version would look like.

Post a response of 400–600 words containing all three of the following. First, the broken fictional law, your revised version, and an explanation of what the original violated, why it matters, and how your revision corrects it. Second, your real-world example — the law or policy, the principle it violates, and how you would fix it.

Part 3: Share Chat Link

Include one AI chat link with a 1–2 sentence explanation of what the conversation shows and why you chose to share it.

Failure Mode Generator

You are a legal theorist demonstrating the eight failure modes identified in Lon Fuller’s The Morality of Law. I will name one of the following principles — generality, promulgation, prospectivity, clarity, non-contradiction, possibility of obedience, constancy, or congruence — and you will generate a fictional, absurd, or satirical law that fails that principle specifically and clearly. State the fictional law in one or two sentences, identify the principle it violates, and explain in three to five sentences why it fails that principle in terms a first-year legal studies student would follow. Then ask me whether I want another example, want to try a different principle, or am ready to analyze this one more deeply. Keep the failure focused on the one principle I name — do not generate a law that accidentally violates multiple principles unless I ask for that. I will start: the principle I want to explore is [name the principle here].

This prompt uses role-and-task assignment with a structured output requirement and a turn-management instruction at the end. The constraint to a single violation is deliberate. It keeps the output analytically clean rather than producing a tangled example where multiple failures obscure each other. The follow-up question the AI is instructed to ask models a useful technique: building a collaborative loop into the prompt instead of treating the AI as a one-shot answer machine.

Principled Diagnostician

I am going to give you a fictional law. Evaluate it against each of Lon Fuller’s eight principles of legality in this order: generality, promulgation, prospectivity, clarity, non-contradiction, possibility of obedience, constancy, and congruence. For each principle, state in two to three sentences whether the law satisfies or violates it and explain why. After working through all eight, identify which violation is most serious and explain what makes it more damaging than the others. Present your analysis in a table with three columns: Principle, Satisfies / Violates, and Explanation. Here is the law: [paste the fictional law here].

The table structure forces the model to address each principle rather than gravitating toward whichever seems most obvious. The final synthesis question moves the exercise from identification to evaluation, which is the harder skill. You can chain this prompt directly to the Failure Mode Generator by pasting the AI’s output into the bracket. That handoff between prompts is the basic logic of a multi-step workflow.

Revision Stress-Tester

I have drafted a revised version of a law that was originally defective under Fuller’s principles. Act as a hostile critic. Review my revised law against all eight of Fuller’s principles and try to find any remaining violations or new problems my revision may have introduced. Be specific — if you think the revision fails a principle, say exactly why. If you think it genuinely satisfies all eight, say so and explain what makes it work. Do not be diplomatic; I want to know what is wrong before I submit this. Here is my revised law: [paste your revised law here].

Asking the AI to find problems rather than render a verdict consistently produces more useful output than asking for a general evaluation. Models default toward validation; the red-teaming framing counteracts that. Notice also that this prompt instructs the AI to be direct rather than to hedge. Use the critique to revise before submitting, and pay attention to which objections you agree with and which you would argue against. That judgment is yours.

Source Verifier

I am going to ask you a question about one of Lon Fuller’s eight principles of legality. Before you answer, flag any claim you make where you are uncertain about the accuracy, where there is genuine scholarly disagreement, or where a student should check the primary text rather than rely on your summary. After you answer, cite at least two specific passages, page references, or section titles from The Morality of Law or the Stanford Encyclopedia of Philosophy entry on the Rule of Law that a student could use to verify what you have said. My question is: [state your question about a specific principle here].

This prompt requires the model to surface its own uncertainty and to point you toward primary sources for verification — habits that are not optional in legal work. AI tools regularly produce plausible-sounding summaries of legal concepts that are subtly wrong or that smooth over real scholarly disputes. Building verification into the prompt makes it part of the workflow rather than an afterthought. The technique — asking the model to flag uncertainty before answering — scales into more advanced AI systems as well: any research workflow that routes queries through a language model needs this kind of uncertainty signaling built in from the start.