Philosophy

Pacisco/Philosophy

Proof versus reasonable belief

Pacisco is intended to help someone reviewing, supporting or attacking a policy/theory/position/etc. come to a reasoned assessment of its validity. The use of Pacisco may not establish a universally acknowledged and  incontrovertible truth.  Reviewers may have reasonable differences of opinion after consulting Pacisco.  However, it is intended that they will have tested the consistency their beliefs and come to a clearer understanding of controversies.

With regard to policy arguments, the original motivation for Pacisco: The notion of truth with regard to complex policy arguments is problematic.  The effect of a policy that has not been implemented must always remain unknown.  Until it is actually tested by implementation, its success or otherwise may only be surmised.  However, the arguments that are advanced in its support or opposition may be determined to be good or bad, and if good may encourage confidence in acceptance or rejection of the policy.  This may still be misplaced; the eventual factors that would determine success or failure of a policy may not even have been acknowledged in arguments made.  However, this is not intended to be a counsel of despair.  Experience tends to confirm that policies that have been subject to intense scrutiny before implementation are more likely to be successful.  Arguments advanced may be tested and those shown to be bad can be dismissed with a greater degree of certainty.

The notions of good and bad with respect to a policy argument may be defined thus;

  1. A good argument is one that has not been shown to be bad.
  2. A bad argument is one that has been shown to be unsound by virtue of its logic or factual content.

The status of a good argument is usually tentative; it may be revised to bad at any time flaws are discovered.

Belief and its revision

At any moment, the reviewer of a policy may hold tentative beliefs about the plausibility of particular arguments.  In Pacisco, it is assumed that these beliefs are not monotonic, i.e. exclusively true or false.  Instead they range continuously between those categorical extremes.  Perhaps logical or mathematical statements may be believed to be incontrovertibly true or false, as may elements of ethical or religious dogma.  Outside of these, arguments and their claims may be believed to be more or less plausible.  In Pacisco, plausibility is represented numerically as values between one (absolute certainty of truth) and zero (absolute certainty of falsity).  A value of 0.5 indicates complete uncertainty or ambivalence.

Pacisco treats arguments as being hierarchically composed of  sub-arguments that support or refute their super-argument.  Revision of the degree of plausibility of a sub-argument should cause revision of that in the super-argument, and so on up to the prime claim.  Explicitly representing plausibility numerically enables this chain of revision to be computed. In Pacisco the calculation is based on Cox’s Theorem; that probability theory may be used as a general logic of plausible inference.  [See Van Horn 2003].  The justification of this approach requires a substantial separate section: however, the reviewer of arguments is not required to accept the result of this calculation.  It is intended merely as a suggestion as to how they might revise their opinion.  At the same time the collective assessment of other reviewers (the average of  individual assessments) is presented for comparison.  The reviewer may assess how far their assessment corresponds with or diverges from the collective.

Some thoughts from Bertrand Russell on revision of belief:

“There are three ways of arriving at an opinion on any subject. The first is to believe what one is told; the second is to disbelieve it; and the third is to examine the matter for oneself. The overwhelming majority of mankind practise the first method; of the remainder, the overwhelming majority practise the second; only an infinitesimal remnant practise the third.
To believe what one is told is the right method for most people in regard to most questions. I believe there is a place called Vladivostok because the atlas says so and because I have met apparently veracious people who assert that they have been there. But if I were engaged in making a survey of eastern Siberia for the Soviet Government, I should have to verify the existence of Vladivostok for myself. Believing what one is told is proper whenever there is a consensus except in matters on which one is a professional expert. In many of the most important questions there is a local but not a world-wide consensus.
To disbelieve what one is told is the method of the rebel and as a general practice has nothing to recommend it. Wisdom is not achieved by refusing to believe that 2 and 2 make 4, or that there is such a place as Vladivostok. When the authorities are unanimous, they are usually right; when they are not, the plain man does well to suspend judgement. A general habit of intellectual rebellion is more foolish than a general habit of intellectual acquiescence, and if it became common it would make civilisation impossible.
It is wise, however, to feel some degree of doubt, greater or less according to circumstances, as regards even universally accepted opinion. Few things seemed more firmly established than the Newtonian theory of gravitation, yet it turned out to need correction. The rational man, in such cases, acts upon the accepted opinion but is willing to give a hearing to anyone who advances serious reasons against it.
Rationality is shown not so much in what you believe as in how you believe it. You are rational if you believe it on evidence and as firmly as the evidence warrants and if, further, your belief leads you to act only in ways which are no obstacle to the discovery of error.
Freedom of opinion is important, since, without it, no generally received error can ever be corrected; therefore no belief should be so firmly held as to lead to persecution of those who reject it. But so long as freedom of opinion is safeguarded, all except professional experts have a better chance of being right if they accept than if they reject the prevalent opinion.”
— Bertrand Russell, Mortals and Others, Bertrand Russell’s American Essays 1931–1935, Vol. II, Essay. 37: What to Believe, p. 454 (24 August 1931)