Universalisability
from 1785 Kant's Groundwork of 1785; restated as a constraint on moral language by R. M. Hare in 1963
State the rule you are actually acting on, imagine everyone in your position following it, and see whether it still works.
- Active
- from 1785
- Where
- Königsberg
- Status
- In use
- Tradition
- Devices and constraints
What it is
Universalisability is a constraint rather than a procedure. It does not tell you what to do; it rules out a class of answers — the ones that depend on your being the exception.
The test has two failure modes and they are different. Some rules cannot be universalised at all, because universal adoption destroys the practice the rule exploits: a rule permitting the false promise cannot be general, since general adoption removes the trust that makes promising possible. Other rules could be universal but could not be willed, because you would refuse them from the other side of the transaction.
How it works
The work is in the first step and almost nobody does it honestly. State the rule you are actually acting on, at the level of description you would accept from someone else — not "park in the bay when the errand is urgent", which is drafted to acquit you.
Then universalise, to everyone in your position rather than everyone alive, and ask two questions: does the practice survive, and would you accept the rule from the other side? A rule failing either fails — not because the consequences are bad but because it is not a rule.
How it is applied
- State the rule you are actually acting on, in terms you would accept from another person
- Imagine it adopted by everyone in your position
- Ask whether the practice the rule relies on survives universal adoption
- Ask whether you could accept the rule from the other side of it
- If either question fails, the rule fails, whatever the outcome in your case
The device applied
Proposition Taking the disabled bay for five minutes while the shop is empty is fine.
- State the rule honestly. Not "park in the bay when nobody needs it" — you cannot know that, and the qualification was added to make the rule pass. The rule you are acting on is: park in the bay when your errand is short and the bay looks unused.
- Universalise it to everyone in your position, which is everyone with a short errand and eyes.
- The practice does not survive. The bay's entire function is being available on arrival to someone who cannot walk from the far end of the car park, and a bay that is free only when nobody with a short errand has noticed it does not perform that function. Universal adoption removes the thing the rule was quietly relying on: that other people are not doing this.
- Now the second question. Take the other side — arriving with a walking frame to find the bay taken by someone who will be five minutes. You would not accept the rule.
- The rule fails both tests, and it fails them before any calculation of how likely it was that someone would arrive.
Therefore The parking is not wrong because harm resulted. It is wrong because the rule permitting it works only while almost nobody follows it, and a rule that requires its own rarity is not a rule you were entitled to act on.
This is universalisability rather than a consequence calculation because the empty car park never mattered. The device tests the rule, not the episode, and it returns a verdict in the case where the harm was nil.
What it is good for
Exposing free riding, which is exactly the class of act that looks harmless one instance at a time. It is also the sharpest available test of special pleading, because the first step forces the description of the act to be one you would accept from anybody else.
Where it fails
The maxim-description problem sinks it if you let it. Any act universalises if the rule is drawn narrowly enough — park in the bay on a Tuesday, in the rain, when the errand concerns a sick relative — and nothing inside the test says at what level to describe the act. Hegel put the charge as empty formalism: the test returns whatever the description put into it.
It also produces rigid answers where nearly everyone thinks flexibility is right. Kant's own application, that you may not lie to the murderer at the door, has been the standing embarrassment of the position since Constant raised it in 1797.
And it is silent about most of life. The test rules rules out; it does not rank the ones that pass.
Common misconceptions
That it means asking what would happen if everybody did it. That is a consequence test and gives different answers — everyone becoming a baker would be a disaster, and baking is not thereby impermissible. The question is whether the rule can be universal, not whether universal adoption would go well.
Where it came from
Kant, in the Groundwork of the Metaphysics of Morals of 1785, as the first formulation of the categorical imperative. The intuition is older and appears in the golden rule, but Kant's version tests the rule rather than the feeling and is a different instrument for that reason.
What became of it
Detached from Kant's system and widely adopted without it. Hare rebuilt it in the twentieth century as a feature of moral language itself — that calling an act wrong commits you to calling relevantly similar acts wrong — which keeps the test and drops the metaphysics.
Lasting influence
It is the reason "would you accept that from me?" counts as an argument rather than a rhetorical question, and it underwrites most contemporary reasoning about fairness, including the veil of ignorance.
Why it still matters
It is the everyday test for whether an exception you are about to make for yourself is one you would grant to others, and it works in ordinary life more reliably than in the seminar cases that are usually used to attack it.
Put another way
A stencil: if the shape cannot be printed across the whole wall, it was never a pattern, only a mark.
Texts
Immanuel Kant, Groundwork of the Metaphysics of Morals
The second section carries the formulations and the four worked examples. The third section is the defence and can be left until later, or forever.
Where to start
Groundwork, second section, and read the four examples first — the false promise is the one where the test does exactly what it claims.
The surprising bit
Kant thought the test was available to anyone without instruction, and said the common human understanding applies it perfectly well without philosophy. The philosophy was for defending it, not for using it.
In their own words
Act only on that maxim which you can will to be universal law
Who held it
Groundwork, second section
Connections
- Relies on
- The thought experiment
- Influenced
- The veil of ignorance, Reflective equilibrium
- Parallel to
- The veil of ignorance
The verdict
The best instrument in ethics for catching the exception you were making for yourself, and defenceless against a sufficiently creative description of your own act.