Tag Archives: Richard Carrier

Hierarchy of Methods

Last time, we considered Richard Carrier’s two common-sense ways of deciding whether a method for finding the truth about anything is reliable:

  • Does it (can it) make predictions that come true?
  • Does it produce a convergent accumulation of consistent results?

It’s almost comical to see how we humans prefer the worst methods and shun the best. Actors and sports stars earn millions for their paid endorsements because they exert a strong and measurable influence on our buying decisions, yet what could be less reliable than the bought-and-paid-for testimony of a non-expert?

At the other end of the spectrum, it is said that popular books on science lose half of their remaining readers with every equation. And those are readers who are presumably interested in science in the first place!

Although I have never paid much heed to celebrities and I love equations, I am shocked to look back on my life and see how unsound my epistemology has been. Even my logic and reason has often taken as its starting point the conclusions of less-reliable methods. I have been far too impressed with expert testimony, choosing to trust people because they shared some of my views. A shared (inbred?) set of views ought to have made me cautious, not trusting. I thought I was using the critical-historical method when in fact I knew next to nothing about what distinguishes a sound historical investigation from hear-say.

With all that in mind, here is Richard Carrier’s list of methods, from best to worst. Quotes are from his book, Sense and Goodness Without God: A Defense of Metaphysical Naturalism, chapter 3. As you go down the list, see if you agree that those near the top have the most “predictive success” and “convergent accumulation of consistent results.”

The method of pure reason by logic and mathematics takes first place. “The logical-mathematical method is supremely successful [because it has] produced the broadest, most complete and most consistent success. Moreover, when a proposition of logic or mathematics is challenged and seriously debated, the most widespread and solid agreement is achieved in comparison with any other method or subject.” Unfortunately, this method is not always available, so we continue with…

The method of science. Science is not perfect, but “we have proven again and again that the results of thorough scientific investigation are more reliable than the results of our own casual observation, producing far more extensive agreement and far more surprising successes, with the most impressive examples of convergent knowledge in history.” When careful science is not possible, the next best thing is…

The method of experience. Although “I saw it with my own eyes” may seem to establish the strongest possible case, Carrier cautions that “if someone comes up with a scientifically or logically well-proven claim that contradicts our direct experience, then we have good reason to believe that our experience is in error, because a single unexamined experience cannot possibly be more trustworthy than a hundred well-analyzed and tested ones.”

The critical-historical method is next. What distinguishes sound, critical-historical investigation from sloppy, biased work is a huge subject in itself. Carrier (an historian himself) says that our skills here are much better than they used to be, but this method still “takes fourth place [because] lacking the ability to observe its object directly, its results are as indirect as its evidence, and by being less direct, [are] less certain.”

Expert testimony is surprisingly far down the list. But when you think about it, expert testimony even at its best is only derived from the more reliable methods above. Sometimes, testimony is given on very thin evidence. As Carrier says, in some fields “we find very little agreement among qualified experts [i.e., little convergent accumulation of consistent results], and a vast influence of ideological bias that is rarely placed under any objective control.”

The method of plausible inference, like expert testimony, starts from the more reliable methods and extrapolates from them, so it must be less reliable. Still, it is sometimes valid to make inferences from a few facts to general principles.

Pure faith has proven to be the least accurate. This is not a slam on any particular religion; it refers to the method of faith in general, not your faith in particular. Many religious people cite evidence from the other methods to support their views, and you may be one of them. Still, Carrier urges us to consider the number of claims based solely on faith have been overturned through history. This ought to make us cautious about faith as a method.

What do you think of this list? Is it as sobering for you as it was for me? Do you aspire to shift your thoughts higher on the list?

Sound Method

In Software Patches for the Brain, I quoted Richard Carrier:

…if you don’t employ a sound method routinely and vigorously, then your entire belief system will be unsteady and imperfect.

Carrier offers two common-sense characteristics of a good method:

  • predictive success and
  • convergent accumulation of consistent results.

These attributes are easy to understand and have obvious merit, yet many of us muddle through life oblivious to them. When our method makes predictions that prove false (or is unable to make an predictions at all), we make excuses. When contradictions arise, we call them truths-in-tension.

Predictive Success

A method that predicts things (especially surprising things) that are later confirmed is more credible than one that cannot make predictions or makes false ones.

The most spectacular example of predictive success may be the discovery of microwave background radiation from the Big Bang. The Big Bang Theory had predicted such a thing, but nobody had observed it. In 1964, Arno Penzias and Robert Woodrow Wilson discovered it by accident as they tried to find the cause of mysterious static on their radio antenna.  Score 1,000 points for the theory. Add 1,000 bonus points for the most understated title of a Nobel-prize-winning paper in history: A Measure of Excess Antenna Temperature at 4080 Megacycles per Second.

Not only was that a victory for the Big Bang Theory; it demonstrated the power of the scientific and logical-mathematical methods.

Convergent Accumulation of Consistent Results

If we find our belief-system spawning more and more contradictions for which we must fashion more and more ad hoc explanations or ugly contortions, that’s bad. On the other hand, if the arrival of additional data confirms our ideas, then we are probably on the right track. That’s especially true if the additional confirmation comes from different directions.

The best example I can find is the evidence for the Theory of Evolution. At one time, I was a creationist. After I read the book, Scientists Confront Creationism, and saw that independent disciplines ranging from paleontology to molecular biology all converged on the same evolutionary conclusion, and creationism could only accommodate these diverse data by twisting all honest thought, I had no choice but to change my view.

Not only did I change my view about creationism, but I was duty-bound to question the validity of the method by which I had come to believe that idea. That was the method of faith, and trust in authority.

These examples show why Dr. Carrier puts the methods of logic, mathematics and science at the top of his list, and ranks methods like faith and expert testimony lower.

More on that next time!