Explication
from 1950 Logical Foundations of Probability
Do not analyse the vague everyday notion — replace it with a precise one that does the same job better, and defend the replacement on those grounds.
- Active
- from 1950
- Where
- Chicago
- Also
- Explikation
- Status
- In use
- Tradition
- Analytic tools
What it is
Explication gives up on capturing what a word means. Carnap's proposal is that where an everyday term is useful but vague, the philosopher's job is to construct a precise replacement — the explicatum — and justify it by what it lets you do.
The replacement need not match the original everywhere. It has to be similar enough to serve the purposes the original served, and it has to be better in the ways precision is better: exact, fruitful, and simple.
How it works
Describe the vague original clearly enough to see what work it does. Propose a precise replacement. Check it against Carnap's four requirements: similar enough to the original, exact, fruitful in the theory it belongs to, and as simple as those allow. Adopt it as a replacement and say plainly that it is one.
The last part matters. Explication is honest about substituting; failing to say so is how a stipulation gets mistaken for a discovery.
The moves, in order
- Describe the vague original and the work it does
- Propose a precise replacement
- Test for similarity to the original in the cases that matter
- Test for exactness, fruitfulness and simplicity
- Adopt it explicitly as a replacement, not as an analysis
The procedure
Proposition The maintenance clause should not say regularly.
- The lease requires the tenant to service the pump regularly. Regularly is doing real work — it rules out never — and it cannot be applied by a stranger.
- The replacement must be similar in the cases that matter: it must catch a tenant who never services the pump and clear a tenant who services it sensibly.
- Proposed: serviced at intervals not exceeding 250 running hours or twelve months, whichever falls first.
- Similarity holds — every case the parties agree about falls the same way under both. Exactness holds — the hour meter and the calendar settle it. Fruitfulness holds — it connects to the manufacturer's schedule and the warranty terms. Simplicity holds — it is one sentence with two numbers.
- It is not what regularly means. Nobody using the word had 250 hours in mind, and the clause is better for the substitution.
Therefore Replace regularly with the interval, and record that this is a replacement rather than an interpretation of what the original clause already required.
This is the whole difference from conceptual analysis. Analysis would ask what regularly means and would get nowhere, because it does not mean anything precise. Explication asks what it is for and builds something that does that.
What it is good for
Building technical vocabulary out of ordinary vocabulary. Every measurable definition in science, standards and legislation is an explication, and most of them are not labelled as such.
Where it fails
It can lose the thing that made the original worth having. Carnap's requirement of similarity is loose, and a replacement can meet it while quietly dropping the cases the original existed for. Strawson's objection is that changing the subject is not the same as solving the problem, and it has never been fully answered.
It also gives the explicator a great deal of power. Since the replacement is justified by its fruitfulness for a theory, the theorist ends up specifying the terms in which their own theory will be assessed.
Common misconceptions
That it is just stipulation. The similarity requirement constrains it: an explicatum that gets the central cases wrong is a bad explication, not merely a different one.
Where it came from
Carnap sets it out in the opening chapter of Logical Foundations of Probability in 1950, having spent years failing to analyse probability and concluding that the everyday concept was two concepts and neither was precise.
What became of it
Increasingly seen as the honest description of what philosophers were doing all along, and as the direct ancestor of conceptual engineering.
Lasting influence
The modern definitions of temperature, computability, validity and probability are all explications, and each replaced a vague predecessor that could not have been analysed.
Why it still matters
High and rising, as conceptual engineering has revived the underlying idea.
Put another way
Replacing a rule of thumb with a gauge. The gauge is not what the thumb meant.
Texts
Rudolf Carnap, Logical Foundations of Probability
Chapter 1 sets out explication and the four requirements. The rest is the worked application to probability.
Where to start
Read Carnap's first chapter, then find an explication in a standard you use at work. There will be one.
The surprising bit
Carnap's response to the failure of analysis was not to try harder but to change the goal, and the move that looked like a retreat has outlasted the programme it was retreating from.
In their own words
The task of explication consists in transforming a given more or less inexact concept into an exact one
Who held it
Alan Turing
Connections
- Relies on
- Conceptual analysis
- Set against
- Ordinary language
- Influenced
- Conceptual engineering
- Parallel to
- Conceptual engineering
The verdict
The most honest thing to do with a vague and useful word, provided you say that you are doing it.