Reason Hell: Authority, Ad-Hominem, and credibility Part 2

Mr Atomic endorses this post
Mr Atomic endorses this post

In Part 1 I discussed how people and credibility are unavoidable when it comes to argument and reason. An individual simply doesn’t have the time or brain power or expertise to do everything themselves and an individual cannot be everywhere at once to witness events as they happen.

We have to, at times, rely on others for
* information another person observed at a specific time and place
* information a person has researched and gather into one place
* analysis, proofs, experiments and other processes that another person has completed
* professional appraisal of facts by known experts
* interpretation of lengthy or technical documents by an expert
* translations of text from another language

Law and medicine are obvious areas in which this reliance on others is of great important but the situation is similar in many areas of human activity. A mathematical proof maybe the paradigm of an objective fact but very few people can actually check rigorously that a proof has been completed without error. We are obliged to trust that claims about mathematics have been checked properly by other mathematicians.

One solution would be to simply trust authority. However, I believe that is an error.

On the other side of the scale we know that our world is filled with people who are less than honest:
* people lie
* people cheat
* people mislead
* people are selective about what they tell you

And of course, sometimes honest people are just plain wrong.

One solution would be to simply mistrust everybody. However, I believe that is also an error.

So we have two kinds of error – trusting authority versus mistrusting everyone. We also have a combined error of mistrusting people society finds disreputable and trusting people that society finds reputable. This combined error leads to a situation in which truth is judged primarily on social position, wealth or political authority. Drugs addicts, criminals, murderers can tell the truth and scientists, police officers and members of royalty can lie

Things are not hopeless though. What we need to consider when dealing with claims made by others is what stands behind those claims. We can’t manually check everything that everybody says but we can look for proxies that indicate how trustworthy a claim might be.

  1. Process matters. Claims by an authority figure (say a scientist) should rest on a process that has been conducted in a valid way. You can’t necessarily check the experimental result in a scientific paper but you can look at the experimental design in some cases.
  2. Accountability matters. Claims made by an authority figure (say an academic of any kind) should have been examined by other experts in the same field. Critiques of their claims will have been made and checked. Those critiques should have been responded too.
  3. Relevant expertise matters. Being an authority on one things does not make a person automatically an authority on another thing.
  4. Consensus matters. Many different authorities in the same discipline coming to the same or similar conclusions independently is significant.
  5. A lack of bias matters. Does an expert have a reason to lie or a motive to ignore relevant evidence? Is there a systemic bias in their field?
  6. Epistemologically efficacy matters. An expert can only know what it is possible for them to know. An expert who makes claims of fact beyond what their field of study could possibly ascertain probably isn’t telling the truth.

Additionally there are reasonable assumptions we can make:

  1. Generally people need a strong reason to lie. There may be come people in the world who lie for little or no reason but most people don’t. People may be deluded or mistaken but generally be relay what they believe to be true because it is easier to do so than lying.
  2. Coherent claims are more likely to be true than incoherent claims.
  3. Multiple lines of evidence matter.
  4. Claims that have stood up to questioning/investigation are more likely to be true.
  5. Reliability matters. A track record of active dishonesty can’t be ignored.

None of these things are guarantees of truth but they provide ways of helping evaluate evidence from other people.

Blog at

%d bloggers like this: