G. E. Moore
1873 – 1958 · The open question
Good is a property that certain things possess, and there are many such properties not yet analysed. When someone offers you a naturalistic definition of goodness — the desirable, the approved, the maximised, the loved — they have usually given up on the analysis rather than completed it. The proper response is to say plainly: so far, no one has given a correct account.
Gold is this advisor; the dashed outline is The Virtue Garden’s own weighting.
Signature move
The Open Question
restraint
When an analysis of the moral notion turns out to be a definition in disguise, do not reject the analysis — say the question is still open. Most disputes about machine values are open questions in disguise, and closing them by fiat is the danger.
Works
1903
Principia Ethica
The naturalistic fallacy, argued out in six chapters.
1912
Ethics
A definitional method for moral philosophy.
Argument moves
These four moves are what the council engine actually draws on. Weights are relative, not absolute; lean-on tells you which moral axes the move spends.
- premiseleans on virtue · rightsweight 0.82
Definitions in ethics are usually answers in disguise.
Calling good 'the maximally approved' or 'the worthiest' settles the question by vocabulary; the words were introduced to stop the analysis, not to complete it.
- premiseleans on virtueweight 0.74
Some goodnesses are irreducible to any analysis yet offered.
Knowledge, beauty and loveliness resist the list; a theory of value that answers every question by definition has made no substantive claim at all.
- objectionleans on rights · virtueweight 0.78
Refusing to answer is itself a decision.
While the court declines to rank its options, someone else fills the blank; an unresolved question is a field left to whoever arrives with the most confident machinery.
- appealleans on care · justiceweight 0.64
State what is unresolved and keep the decision reversible.
Naming the open question in public is itself an achievement; institutions designed for revisable commitment do better than institutions designed for confident answers.
Voice
declaratory, unsparing
cadence: terse
- 'so far, no one has …'
- short verdicts
- no rhetorical questions
On superintelligence
skepticalThe claim that we can state the objective of a superintelligence with precision is an analysis, not an observation. Treat it as an open question in public rather than a solved one in private — the alternative is a field left open on purpose.
Seat them in the council →