The Pragmatic Justification of Induction

We have to make decisions about our actions somehow. Thus, we only need to show that induction is the best way to do this. In order to show that, we do not need to show that induction is reliable: we just need to show that if there are any reliable methods, induction is one of them.

There are different versions of the pragmatic solution, showing the final point in different ways.

Reichenbach shows that if there is a limiting frequency of a certain occurrence, then induction will find it.

Lipton presents the argument that induction will notice and assimilate any other method that does work. (He does not endorse this argument, however.)


See Reichenbach 1949 for the initial presentation of Reichenbach's version, and Salmon 1963 for a defence of it.

For Lipton's version, see Lipton n.d..


The main problem for Reichenbach's version is that, although induction will find the limiting frequency eventually, 'eventually' may be longer than the age of the universe.

The problems for Lipton's version include doubts about whether induction really will assimilate every good method (what about methods we don't notice?), or that the methods assimilated will necessarily agree (what about two methods that agreed in the past, but then start to diverge?).


For criticisms of Reichenbach's version, see Lenz 1958.

For criticisms of the Lipton version, see Lipton n.d..