David Hume And The Justification Of Inductive Inference

The original problem of induction concerns the justification of inductive inference.
Hume maintains that it is the past regularity which establishes a habit that makes prediction happen. Goodman thinks Hume grasp the essence of the problem but also points out that not all regularity can form a habit to guide the prediction. The regularity which Hume refers to is only the generalization of evidence statement of something, not everything. For instance, a given piece of copper conducts electricity may convince people that other copper conduct electricity, however, the man in the room is the third son does not convince us that other men in the room are also the third son.
Goodman states that the real problem in explaining “how induction happens” is not the difficulty of justifying the forms of inductive inference, instead it is how we can distinguish valid and invalid predictions. He argues that Hume only focuses on regularities but fail to distinguish the projectible from non-projectible regularities.

The process of justification looks like a circularity that rules and inference are being brought into agreement by making mutual adjustments with each other. These rules also apply to the justification of induction. Predictions are justified if they conform to the valid canons of induction whereas the canons are valid if they codify the accepted inductive practice, but the what is the valid canon remains in