Carl hempel

Carl Hempel on Stanford Encyclopedia of Philosophy

The role of theories in science is to explain phenomena and theories derive their explanatory power from a series of general statements from which stem the description of specific instances.

General statements

These general statements are what we commonly call a scientific law.

Of course, to achieve their explanatory value, they must have a set of features:

[laws] are […] statements of universal form. Broadly speaking, a statement of this kind sserts a uniform connection between different empirical phenomena or between different aspects of empirical phenomena. It is a statement to the effect that whenever and wherever conditions of a specified F occur, then so will, without exception, certain conditions of another kind, G.

Laws and Their Role etc., Aspects, p.54

Features of scientific laws:

  • Statements are of a universal form
  • yet, they’re relevant to what I’m seeing here and now

They allow to find out what is happening at a given moment.
In general, whenever F, then G.

Every time we try to explain some event, we need to refer to ha general explanatory law. Once we have a law, we have an explanation.

The Deductive-nomological model

The formalization of the previous idea is formalized as the Deductive-nomological model (The term ‘nomological’, from the Greek word νόμος (law) means ‘pertaining to laws of nature’.).

Premises are defined as the explanans, and they are formed by a combination of:

  • at least one general law
  • multiple initial conditions

from this, we are then able to infer the explanandum, that is what needs to be explained; it does not specifically refer to the actual fact in the world, but it is a sentence which describes it.

The explanation consist in deriving an account of the explanandum from the existence of at least one law which seems to describe the case in question.

Such model is deductive because, if the law is true, then deductively it will necessary lead to a correct explanation.
On the assumption that we’re using a correct law, then we’re sure the logical connection is valuable.
The explanation is the statement, it’s the connection

Nevertheless, how can we state the validity of a law?
How do we account for the correlation a law supports?

Accounting for the laws

There are two ways of accounting for the laws:


Picking the right level of universality for statements.

  • Accidental generalizations: general statements, yet limited to the contingent observation in question.
  • On the opposite, there needs to be general statements which are always valid (e.g. all gasses expand when heated under constant pressure)

Can universality be tested?

Substitute the particular observation in an instance in a if-then form: if a universal statement is a good candidate as a scientific law, it should be able to work.

These statements are conterfactuals: they’re not actual events, but they’re purely hypothetical, not empirically performed. Yet, they’re needed to prove the validity of a statement as a law.
They are a test for genuine universality. A test for the capacity of the statements.

For example:
p1: all apples in this basket are red
p2: all gases expand when heated
counterfactual of p1: if this apple were put in this basket, then it would be red counterfactual of p2: if the oxygen in this sylinder had been heated under constant pressure, then it would have expanded

Second way

Laws are ceteris paribus, all things being equal:
Laws hold the truth when they’re able to keep away elements which are beyond the scope of the law.

The example of Galileo laws of motion: they’re true under highly idealized conditions. The general law idealizes the conditions in which it should be applied.

This led some philosophers of science to question the actual universality of laws.

Next topic: Nancy Cartwright



edit this page

The following is a graph containing all the notes and every topic in the website