The open question
from 1903 Principia Ethica
Whatever you identify goodness with, it always remains sensible to ask whether that thing is actually good — and that it remains sensible is the argument.
- Active
- from 1903
- Where
- Cambridge
- Status
- In use
- Tradition
- Devices and constraints
What it is
Moore's argument targets any proposed identification of a value with a natural property. Suppose goodness is pleasure. Then asking whether pleasure is good should be like asking whether a bachelor is unmarried — closed, trivial, not a real question.
But it is not. It remains a sensible thing to ask, and someone who asks it has not misunderstood the words. Moore concludes that the identification is false, and that this holds for every candidate anyone proposes.
How it works
Take the proposed identification of the value with some property. Form the question: is what has that property actually good? Ask whether a competent speaker could sincerely and intelligibly ask it. If they could, the question is open and the identification fails.
The test is whether the question is closed, not whether the answer is obvious. Plenty of open questions have obvious answers.
The form of the inference
- It is proposed that goodness is identical to property N
- If so, asking whether N-things are good would be a closed question
- A competent speaker can intelligibly ask whether N-things are good
- Therefore the question is open and the identification fails
The argument
Proposition The efficiency criterion cannot be the whole of the selection criterion.
- The procurement framework identifies the best tender with the one offering the lowest whole-of-life cost per unit of service.
- If that identification held, asking whether the lowest whole-of-life cost tender is the best tender would be a closed question — like asking whether the cheapest is the cheapest.
- But the panel asked exactly that question about the winning waste tender, and nobody at the table thought they had misunderstood the framework or were speaking nonsense.
- What they were asking was whether a tender that achieves its cost by hauling to a facility two hundred kilometres away, with a subcontractor the council cannot audit, is the best one — and that question is intelligible.
- So lowest whole-of-life cost is not what best means in the framework, whatever the framework says.
Therefore The framework's definition does not capture what the panel is actually assessing, and the criteria it leaves out have to be written down or they will keep being applied without being stated.
Nothing here shows the cost criterion is wrong or that the panel's misgivings are justified. It shows that the definition is not doing the work the framework claims, which is exactly the force of Moore's argument and exactly its limit.
What it is good for
Testing any proposed definition of an evaluative term — best, fair, reasonable, adequate. It shows quickly when a definition has captured a criterion rather than the concept.
Where it fails
It proves too much. The same test defeats every correct analysis, since a competent speaker can intelligibly ask whether a bachelor is an unmarried man if the analysis is not obvious to them — which is the paradox of analysis, and it means the openness of a question might show only that the identification is informative rather than false.
It was also undermined by the semantics of the 1970s. Water is H2O is a true identity and asking whether water is H2O is perfectly open, because the identity is discovered rather than known by competent speakers. If value could be identified with a natural property in that way, Moore's test would not detect it.
And it relies entirely on a judgement about what sounds intelligible, which is exactly the kind of evidence experimental philosophy has found least stable.
Common misconceptions
That the naturalistic fallacy is the same as the is-ought gap. They are related and distinct: Hume's point is about inference from premises to conclusion, Moore's is about the identification of a property with another property.
Where it came from
Moore's Principia Ethica of 1903, which opens twentieth-century moral philosophy and sets its agenda for fifty years.
What became of it
Enormously influential, then substantially deflated by the paradox of analysis and by Kripke-Putnam semantics, and still routinely used as a quick test.
Lasting influence
It produced non-naturalism, emotivism and much of metaethics, all of which are responses to it.
Why it still matters
Its metaethical force is contested; its use as a test of definitions is undiminished.
Put another way
A lock that opens on every key, including the right one.
Texts
G. E. Moore, Principia Ethica
Chapter 1 contains the argument. Short, confident, and the foundation of a great deal that followed.
Where to start
Apply it to any definition of best or fair in a document you work with. It rarely survives.
The surprising bit
The argument that dominated moral philosophy for half a century was undone by the observation that water is H2O is also an open question.
In their own words
Whatever definition be offered, it may be always asked, with significance, of the complex so defined, whether it is itself good
Who held it
the paradox of analysis
Connections
- Relies on
- Conceptual analysis, The counterexample
- Parallel to
- The is-ought gap
The verdict
An unreliable proof and a good first test of any definition of an evaluative word.