5. What Part II Will Examine

Recursive examination may leave a conviction on firmer ground.

More of what it depends on has been exposed and examined, while the conviction remains convincing. We may be less exposed to having something we rely on overturned by a difficulty we could already have brought into view.

That matters when other judgments, decisions, or commitments have been built on the conviction.

But firmer ground does not yet give us a definite criterion for ordering different convictions as better or worse. Firmer ground describes a change in the same conviction. It is, in a sense, better situated than before.

Other convictions can undergo the same kind of examination and also come to stand on firmer ground.

And a conviction on firmer ground can still be false. It can still be inconsistent with other convictions, depend on a poor representation, or leave something important unexamined.

So firmer ground does not by itself give us a single scale on which different convictions can be ordered as better or worse.

Firmer ground differs from the criteria we ordinarily use to compare convictions. We may compare convictions by truth, consistency, evidential support, predictive success, usefulness, or something else.

Once the criterion tells us what is to count in the comparison, we can ask how the convictions fare by that measure. The result may rank one above another, leave them tied, or fail to order them at all.

But the problem from Chapter 2 remains: different criteria can produce different assessments, and Conviction Formation Theory cannot settle which should govern merely by selecting one in advance. The criteria themselves remain within conviction formation and open to examination.

Part II will thus examine forms of inquiry in which, once enough conditions are fixed, some consequences are no longer independently movable.

If we can reconstruct where the constraint within a formal system comes from, and what gives the results of applying that system to a particular problem convincing force, we may learn more precisely what recursive examination can add, under what conditions it leaves us on firmer ground, and whether it gives us any further basis for calling one conviction better than another.

Part II will examine logic, probability, and causal methods as examples of such strong constraint.

Next essay: Sources Of Logical Constraint