Conceptual analysis
from 1903 Moore's Principia Ethica is the usual starting point
Take a concept apart into the conditions that make it up, and test the result against every case you can think of until it admits exactly what it should.
- Active
- from 1903
- Where
- Cambridge
- Status
- In use
- Tradition
- Analytic tools
What it is
Conceptual analysis takes a concept people use confidently and tries to state what it amounts to: the conditions under which it applies, stated precisely enough that anyone could apply it the same way.
The test is always the same. Does the analysis admit cases it should exclude, and does it exclude cases it should admit? A proposed analysis lives or dies on that, and the person doing the analysing is expected to hunt for the cases that would kill it.
How it works
State the concept. Propose conditions. Look for a case satisfying the conditions that is not an instance of the concept, and for an instance of the concept that fails the conditions. Repair or abandon. Repeat.
The distinctive discipline is that you attack your own proposal. An analysis its author has not tried to break has not been analysed.
The moves, in order
- State the concept to be analysed
- Propose conditions for its application
- Seek a case meeting the conditions that is not an instance
- Seek an instance that fails the conditions
- Repair or abandon, and repeat until neither case can be found
The procedure
Proposition The applicant is not a resident for the purposes of the rebate.
- The first proposal: a resident is someone whose name is on the electricity account for the property. This admits too much — a landlord who keeps the account in his own name between tenancies would qualify while living elsewhere.
- Second proposal: a resident is someone who sleeps at the property most nights of the year. This excludes too much — it would disqualify a shearer away for eight months whose only home is the property.
- Third proposal: a resident is someone who has no other place they treat as home, and who occupies the property when not away for work or care.
- This admits the shearer and excludes the landlord, and no further case has been found that it gets wrong.
- The applicant maintains a house in another town where his family live and which he returns to fortnightly.
Therefore He has another place he treats as home, so on the analysis that survives testing he is not a resident.
The two failed proposals are not wasted work. Each one narrowed the target, and the surviving analysis is shaped by exactly the cases that killed its predecessors.
What it is good for
Any term that must be applied consistently by different people to cases the drafter never saw. It is the everyday work of legislation, contracts, standards and eligibility rules, whatever it is called in those settings.
Where it fails
Its central assumption is that concepts have definite boundaries waiting to be found, and Wittgenstein's argument is that most do not — that *game* has no set of conditions common to all games, only overlapping resemblances. If he is right, a great deal of analysis is looking for something that is not there.
Gettier's 1963 paper did comparable damage from inside. Three pages showed that the analysis of knowledge philosophers had worked with for two thousand years admitted cases it should have excluded, and sixty years of increasingly baroque repairs have not produced a version that survives. That looked at first like a hard problem and increasingly looks like evidence against the method.
Common misconceptions
That it is about words. It is about the concept, and the point of the case testing is that the analysis has to match what people actually pick out, not what a dictionary says.
Where it came from
G. E. Moore and Bertrand Russell at Cambridge in the first decade of the twentieth century, taking apart the idealist positions they had grown up in.
What became of it
The dominant method of analytic philosophy for sixty years, badly shaken by Gettier and by Quine's attack on the analytic-synthetic distinction, and now practised more modestly alongside conceptual engineering.
Lasting influence
It gave analytic philosophy its name and its manner, and its habits are visible in every well-drafted definition in law and standards work.
Why it still matters
High in practice, contested in theory.
Put another way
Cutting a key by testing it in the lock, filing, and testing again.
Texts
Edmund Gettier, Is Justified True Belief Knowledge?
Three pages. The most efficient demolition in the history of the subject, and the clearest demonstration of how the method works even as it breaks something.
Where to start
Read Gettier first, then try to repair the analysis yourself. Failing at it teaches the method faster than succeeding would.
The surprising bit
A three-page paper by an unknown philosopher in 1963 destroyed an analysis that had stood since Plato, and the profession has not managed to replace it.
In their own words
It is impossible that anyone should fail to know what this property is, and yet should be able to think of it
Who held it
Edmund Gettier
Connections
- Relies on
- The counterexample, Necessary and sufficient
- Set against
- Conceptual engineering
- Influenced
- The method of cases, Explication
- Parallel to
- Genus and differentia
The verdict
The best method there is for making a term usable by strangers, resting on an assumption about concepts that may not hold.