I. The Ladder Essay 3
Rules That Travel
Principle thinking
There is a sign in a school corridor that says walk, do not run. Nobody argues with it. It is not a law and it is not advice. It is just what happens here, and if you have spent any time in a building with children in it, you can probably guess what happened before the sign went up. Something went wrong. Somebody was hurt, or nearly. And then a group of tired people made a rule out of one bad afternoon.
Now make the sign do something. Take the rule away from the corridor and hold it in your hand. It says: when you are in a corridor, in this building, with people in it, moving at speed is not acceptable. You can now apply it somewhere else. Apply it to a staircase. Apply it to a car park. Apply it to a fire exit. The rule travels. It keeps working as long as the thing it was made for keeps being true, and the thing it was made for is that people in corridors collide with other people in corridors.
A rule that travels is an abstraction that has been given authority over action. That is the whole idea, and it is the reason principles are both the most useful and the most dangerous things a person can hold. Useful because you cannot consult every corridor and every staircase and every new situation in person. Dangerous because the authority travels further than the evidence does.
Four words that are not the same word
"Rule" and "law" and "principle" and "heuristic" get used interchangeably, and they are four different tools with four different failure modes.
A law is a rule with an enforcement mechanism behind it. It is not asking you to agree with it. It says this is what happens here, and there are consequences. Laws are narrow, specific, and owned by somebody. If you do not know who enforces a law, you do not know how seriously to take it.
A rule is a decision about what to do, made in advance, by someone with the standing to make it. "Walk, do not run" is a rule. Rules are usually about recurring situations and they are usually about a specific context. A rule is not trying to be right in general. It is trying to be settled for here.
A principle is a rule that has been made to travel. It has been stripped of its original situation and given a justification, so that it can be applied to cases that are not in front of you. "Treat people with respect" is a principle. "A rule" in the sense of a school sign is not. The difference is not the strength of the instruction. It is the reach.
A heuristic is a rule of thumb, used because it is usually right and cheap, not because it is always right. "The first thing you hear about a problem is probably the real problem" is a heuristic. It fails sometimes. You use it anyway because the alternative is to think for a long time about every problem, and most problems do not deserve that.
The four are ordered by how much they promise and how much they can let you down. A law promises nothing and can hurt you. A rule promises a settled answer. A principle promises to generalise, and fails when it generalises past its evidence. A heuristic promises only to be usually right, and is honest about that. The vocabulary is worth keeping because each word tells you how much to trust the thing and how far to carry it.
A workshop makes kitchenware. Two people, a wheel, a kiln, and a list of orders. The rule they work by is: never promise a date you have not already seen a finished piece pass through the kiln. It was not invented. It came out of a fortnight in which they took an order for a wedding, told the couple the dishes would be ready in time, and the kiln cycle ran long, and the wedding happened anyway. The couple were not difficult. The order was not unusual. The rule was simply a guess dressed as a promise.
Now apply the rule to a new case. A restaurant calls and wants one hundred plates by a certain date, for a large event, and offers good money for the rush. The workshop has the wheel time. The kiln will not fit the schedule. Under the rule, the answer is no, or a smaller order, or a later date. The rule has done exactly what it was made to do: it has refused to promise a date the workshop has not seen a finished piece pass through the kiln. Note that the rule says nothing about whether the order is worth doing. It only constrains one thing, the promise. That is what a good principle does. It answers a question, and only that question.
Where principles come from
Every principle starts as a case. Somebody met a situation, looked at it closely, and noticed a pattern. Then they said the pattern out loud in a form that no longer mentions the situation. That is distillation, and it is the join between the tangible level and the principle level.
This is why the concrete case matters, even for a rule that will end up applying everywhere. A principle distilled from two cases is a guess with confidence. A principle distilled from two hundred cases, in all their messy variety, has a claim to be a general rule. The "walk, do not run" sign was not written by someone who thought about running in general. It was written by someone who had watched what happens when a corridor fills up and somebody moves fast through it. The reason the sign is well designed is not that it is clever. It is that somebody was there.
And this is where principles fail hardest, because you can skip the distillation. You can receive a principle as a finished object, an authority handed down, with no visible cases behind it. That is what dogma is. Dogma is a principle with the cases removed. It still carries the authority. It has just lost the evidence that earned it, and nobody can check it any more.
Two principles can disagree
Here is the part that gets skipped in most accounts. Principles do not agree with each other. A principle is an abstraction given authority over action, and when two of them have authority over the same action, they can pull in opposite directions, and no amount of reasoning about either one alone will resolve it.
Be honest and be kind are the standard example, and they are standard because they are real. A student has written badly and also revealed something private. Telling them the truth about the writing is honest and unkind. Softening it until it is not a criticism is kind and dishonest. The word "honest" does not resolve this, and neither does the word "kind". Both are principles, both are defensible, and together they are in genuine conflict.
What if the school had a rule that said teachers must give detailed written feedback on every piece of student work? The rule looks neutral. It has no moral content at all. Now change one condition: the deadline. The rule still says feedback on every piece. The teacher's evening is now gone. What fills the space the evening used to hold is a set of shorter, blunter comments, and the students who most needed the detailed reading are exactly the ones whose teacher had the least time. The rule did not change and the outcomes did, because the rule was one of several constraints competing for the same resource.
Now change a different condition: what if the rule only applied to students who had asked for feedback? The count of pieces to be read drops, the depth per piece can rise, and the students who never ask remain unhelped. Same rule, different scope, different results. The scope of a principle is part of the principle, not a detail about it.
"A good principle has no exceptions." This sounds like intellectual strength. It is usually a sign that the principle has been used as a weapon. A principle with no exceptions is not a principle, it is a dogma, and dogma is a principle whose cases have been deleted so that it can never be checked.
The opposite misconception is more dangerous, because it sounds humble: "there are no rules, it is all context." This is a way of saying that the level you happen to be working at is the only level that matters. It is right that context matters. It is wrong that context is the only thing. A person who never uses a principle will spend their life rediscovering the same lessons one case at a time, and will never be able to prepare for a case they have not met.
Scope is the part people forget
Every principle has a scope, and the scope is where the trouble is. The scope is the set of cases the principle was made for. Outside the scope, the principle can still be applied, and it may be wrong, and you will have no way of knowing from inside the application that you are outside.
"Walk, do not run" has a scope. Corridors, with people in them, where collisions hurt. Outside that scope it becomes strange. On an empty playing field at a sports day, running is the point. On a fire evacuation route, running is the point too, and a sign saying walk would be actively dangerous. Nothing has changed about running. What has changed is whether the people in the space collide, and whether the space is designed for people to move quickly through it.
Edge cases live outside the scope, and they are not a problem to be solved. They are the price of having a scope at all. A rule with no edge cases has no scope, and a rule with no scope is a slogan.
One practical consequence follows. When you find yourself arguing hard against a rule, check whether you are inside its scope. If you are inside, you are probably wrong and the rule is probably right. If you are outside, the rule might be right in its own place and simply not applicable to yours, and arguing about whether it is a good rule in general is a way of avoiding the question of whether it fits here.
Dogma
Dogma is what happens when a principle outlives its scope and loses its cases. It has three symptoms. It is quoted rather than applied. It is defended by pointing to its authority rather than to the situations it was made for. And it cannot be revised, because the people who would revise it were never given the cases that would let them see the need.
Dogma is not always stupid. Sometimes a principle that was right for a decade stops being right because the world moved, and the people who built it into dogma were sincere. A rule about phone calls in a school corridor was sensible in a building with one corridor and no classrooms in use. It is absurd in a building where every classroom has a phone on the desk. The rule did not change. The scope did, silently, by everything else moving.
The defence against dogma is simple and unfashionable. Keep the cases. When a principle is taught or adopted, keep the situation it came from. Ask what it was made for. If nobody can say, it has already become dogma, and the only honest response is to treat it as a claim to be tested rather than a rule to be obeyed.
There is a small habit that catches dogma early. When you catch yourself explaining a rule by pointing at the rule, stop. Say instead what would have to be true for the rule to be right, and then check whether it is. "We always do this because it is the policy" is the voice of dogma. "We do this because a pupil was hurt last winter, and the corridor was full" is the voice of a principle that still has its cases. The first cannot be questioned without offending somebody. The second can be tested, and when the world changes, it can be updated by pointing at the change.
Keeping the cases also makes a principle teachable. A rule handed over as a rule has to be obeyed on trust, and obedience on trust decays fast. A rule handed over with its story can be reasoned about, argued with, and improved. The students in a studio who have heard why the kiln is never rushed will watch the kiln themselves, and the watching is what keeps the rule alive long after the person who made it has moved on.
Limits
A principle cannot tell you whether it applies to the case in front of you. It can tell you what to do if it does. The judgement about applicability is a separate act, and it happens at the level above. This is why experts seem to break rules constantly, and why novices who break rules seem reckless. The novice breaks a rule without noticing they were in it. The expert breaks it after checking, and can say why.
A principle also cannot be derived without cases, and cases cost time. The pressure to produce a principle quickly is real, and most principles in circulation have been distilled from far fewer cases than their confident tone suggests. Ask how many, and how varied. If the answer is two, you are looking at a rule of thumb pretending to be a law.
Stuck
You are stuck at the principle level when you start applying rules to cases where you have not looked. The tell is a confident answer with no picture in it. Somebody asks why the class did badly and you say "because the baseline was too high", and you cannot say what the class actually did. The principle has arrived before the case.
The way out is to reverse the order. Go and look at the actual work, the actual queue, the actual conversation. Write down what happened in particulars, with no rule words in the description at all. Then, and only then, ask which rule fits. If the rule you reach for does not explain what you saw, the rule is wrong, not the seeing.
Sources
- Heuristics and their biases are well documented in the work of Daniel Kahneman and Amos Tversky on judgment under uncertainty.
- The distinction between rules and principles used here is the author's own working definition, not a borrowed taxonomy.
- All examples are hypothetical or ordinary operational situations. No specific incidents or studies are cited.