Logical Strength 
A summary of Appendix 2 from the revised (1983) edition
The Retreat to Commitment

by William W Bartley
*  ESSAY *

1.The problem of demarcation reconsidered

The kind of demarcation that Bartley is talking about here is the demarcation between good and bad ideas, or between good practice and bad practice.

This is much wider than the usual concerns of philosophers, to demarcate science from non-science, rational belief from irrational beliefs and meaningful statements from nonsense.

2.Demarcation and justification

Coming to grips with the choice between one idea or another, Bartley wrote:

"In a simpler world, one might solve such problems without any explicit recourse to philosophy...For example, I might simply ask my friend Harry which to choose. Or I might flip a coin".

In the less simple world of philosophy there is a desire for justification, that is, we want to justify our ideas, or, using other words, verify, confirm, make firm, probabilify, validate, vindicate, make certain, authorise, defend.  This desire for justification, in its various forms and manifestations, causes us to participate what Bartley has called the "justificationist metacontext".

Bartley suggests that justification (or whatever it is called) involves the following:

a)an authority or criterion for justification.
b)the idea that the goodness or badness of an idea can be determined by reducing it to, or deriving it from, one or other of the authorities or criteria.

3.The justificationist pattern of demarcation

Bartley lists some good traits and bad traits in the Western rationalist tradition.

true   false
probable   improbable
clear and distinct     unclear and indistinct
demonstrable by reasonundemonstrable by reason
empirical          unempirical
verifiable          unverifiable
meaningful       meaningless
scientific           unscientific

Truth is the most prized trait but also the hardest to pin down. Various of the others, singly and in various combinations, have been used as the cardinal criteria of value at different times and places.

4.Problems of logical strength

The problem of logical strength arises when it turns out that justification cannot be achieved. This happens when the statement or policy under analysis cannot in fact (or in logic) be reduced to or derived from the chosen authority.

Bartley points out that traditionally, authorities or criteria have been postulated to solve problems (especially the problem of justification) and it is not usually anticipated that problems are going to arise, so long as the correct criterion is used (hence the bitter disputes between those who use different criteria).

5.Logical strength: An elementary lesson

Bartley explains with some simple examples of logical deduction how the problem of logical strength arises from trying to establish a conclusion of an argument that is stronger than the premises.

For example from the premises

The earth is a planet and moves in an ellipse
Saturn is a planet and moves in an ellipse

We cannot validly deduce the conclusion:  All planets move in ellipses

6.The traditional problems of epistemology as a problem of logical strength

Most modern skeptics and rationalists regard science, based on evidence, as the highest form of knowledge. The problem of logical strength arises here in the form of the problem of induction, that is, the problem of establishing laws or lawlike generalisations (all ravens are black) based on a finite number of observations (of black ravens).

Attempts to overcome this problem, for example Bertrand Russell's lifelong efforts, do not confront or question the underlying structural problem, namely the demand for justification. One approach (used by Russell) is to try to prop up the authority (empirical evidence) by supplementary principles, such as a principle of induction. Another is to abandon the logical issues and simply set out to describe the language of science or the language games of science or to conduct anthropological studies of scientists at work.

7.Turning the tables: nonjustificational criticism

An alternative to the justificationist approach is to simply regard knowledge, or moral principles as conjectural. They are tested, rather than justified, by evidence. The idea that all planets move in elipses is tested by making observations of moving planets. The idea that all ravens are black is tested by observing ravens (if it matters).

We aim to form critical preferences, not justified beliefs, in defiance of Alfred Ayer.

The logic of this situation makes use of the power of deduction and the capacity for a weak statement (an observation of a single raven) to refute a stronger statement (the proposition that all ravens are black).

Bartley claims:

a)This solves the problem of logical strength which will bring undone any attempt to provide strong justification (or certainty).
b)This is an account of empirical criticism, as applied by Popper with this falsification criterion of demarcation.
c)There is no authority, eliminating the first of the two requirements for justification. (Incidentally, this eliminates the need to quarrel over rival authorities).
d)The problem of induction has gone.

8.How other problems are resolved: Realism

Bartley listed a number of problems that have proved to be insoluble within the traditional framework of philosophy - he took these from Russell's The Problems of Philosophy (1912).

These problems include the existence of the external world, the continued existence of the personal self, the existence of other minds, the uniformity of nature, the existence of the past, the reality of time, the existence of matter.

None of these things can be justified in the strong sense (however absurd that may seem) because the attempt falls down every time on the problem of logical strength. However we can form critical preferences one way or the other in the light of evidence and arguments, and we can demand that disbelievers in realism, the existence of matter etc provide alternative explanations for the things that we experience around us that are more resistant to criticism than our commonsense or realist views.

9.Factual information and moral claims

Bartley examined the doctrine of the dualism of facts and values which has been asserted by Hume, Moore and Popper. The doctrine is usually advanced to argue that morals cannot be derived from facts, cannot be justified by facts, and so are beyond rational argument. Bartley argues that this understates the role of facts in moral discourse, even allowing for the dualism.  Factual considerations arise in at least two ways, first, moral exhortations (for example to establish socialism to even out personal wealth) may fail on the grounds that the proposed policy will not work (as a matter of fact). Second, real-world moral options usually come down to the lesser of evils and it is helpful to have some realistic idea of the likely outcome of the options.

10.Evolution, ecology and demarcation

In conclusion, Bartley draws some comparisons and contrasts between the growth of knowledge and biological evolution.