On when to trust epistemic claims

Trusting the word of others is necessary to expand knowledge beyond perception, e.g. we will never know what it is like to be another person (see Nagel’s Bat Argument)

Epistemic Authority with Anand Vaidya

YouTube talk

How do we distinguish between appeals to authority that are rationally/ethically problematic and ones that are morally and rationally wholesome and skillful?

Question about demarcation of appeals to authority between good and bad turns into a much larger question that goes into the philosophy of mind and philosophy of neuroscience, the nature of the self, how that is created, and how that thing epistemically gains and accesses information in the first place through retention

Two concepts and distinctions

  1. Epistemic Culture: what do members of the culture take to be valid sources of knowledge
  2. Epistemic Self: an individual epistemic agent within a culture might have their own idiosyncratic belief about how the source work, how they’re weighted, etc.

Knowledge Sources:

  1. Testimony
  2. Perception
  3. Inference

Shared Epistemic Culture

Throwing something from one epistemic culture that’s disjoint from another epistemic culture cannot lead to reconciliation without a shared epistemic culture (see also: collaborative thinking).

Dogmatism is when one cannot understand differences in epistemic cultures

Critical Thinking

  • Overriding defeater: given evidence presented, this piece of evidence is a countering piece of evidence that overrides the argument for what they’re saying
  • Undermining/undercutting defeater: some kind of source is fundamentally wrong and nothing comes out of it

Is it worth to assess overriding testimonial information from authorities? Is critical thinking good?

Michael Huemer: you shouldn’t! Critically thinking is epistemically irresponsible. Anand’s counter-argument: you always need to critically evaluate at least the following

  1. who are the authorities?
  2. are they currently properly performing their authority?
  3. what does their authority amount to, within their expertise?

In being an expert and giving expert testimony means that we want the expertise that the person has to be spoken for and performed — we want an expert performance. This can be suspect when the information upon which the expert is drawing is ambiguous. Skepticism can also start to creep in because of the bureaucratic complexity through which the message is being delivered

Argumentation

Two views on argumentation

  • Epistemic: trading of assertions has a way of leading to a better epistemic position either for both parties or for the conjunction of the two parties (positive sum)
  • Non-epistemic: trying to pull one side to another (zero sum)

Not only an issue of trust but also psychological exhaustion. One of the requirements of any epistemic enterprise is the use of attention. Information that puts its receivers into an almost constant state of some kind of cognitive dissonance negative impacts their ability to attend and assess the information. See: salience principles of democracy

One of the main components of collective action is collective belief. We want everyone to do but that requires everyone to believe that doing is right

Evaluating authority

Achieving consensus is difficult, especially among a wide net of epistemic agents. In fact, without formal processes like voting or delegation, just peer-based discussion leads to a factorial explosion in time-taken to reach consensus.

What about a “best bet view”: if the weighted average of authorities who meet these tests say something, we should trust that it’s true (this depends on epistemic culture and epistemic self)

Thus in most cases, rough consensus or trusting delegates (e.g. authorities like the CDC) to not be biased and accurately evaluate other authorities (e.g. scientists and pharmaceutical companies responsible for inventing/deploying the vaccine) is usually suitable when balancing tradeoffs between speed (for example when timeliness during a pandemic is important) and correctness (making sure vaccines are safe to the general public).

Increased diversity (of critical thinking skills) leads to a decrease in the probability of the whole group being accurate in evaluating whether or not someone is an authority. You can fool some people some of the time but you can’t fool all the people all of the time

Listen to voices not being heard and figuring out how to repackage what they’re saying in a way that is significant — it is a unique skill to be trained as a cross-cultural philosopher. These philosophers can then teach experts to communicate better. Perhaps this is the role of knowledge distillers like teachers and those who specialize specifically in pedagogy?

See also: testimony

Experts: Which Ones Should You Trust?

by Alvin I. Goldman (2001)

Central Question: the Novice/2-Expert Problem

  • An expert
    • Knows a lot about the domain (first order material)
    • Knows about the literature concerning the domain (second order material)
    • Able to draw on this knowledge to produce answers about the domain
  • Two experts disagree, as a novice, which expert do you trust as more credible? 5 kinds of evidence
    1. Arguments presented by the contending to support their own views and critique their rival’s views
    2. Agreement from additional putative experts on one side or other of the subject in question
      1. Counterpoint: Copernican heliocentrism, most ‘experts’ believed that the Earth was the middle of the solar system, not the sun
      2. Goldman steel-mans his argument through Bayesian analysis, experts should think for themselves and should be ‘conditionally independent’ — they will not invariably parrot the claims of others when those claims are wrong
      3. What about discoverability of experts? If a tree falls but nobody hears it, did it still happen?
    3. Appraisals by “meta-experts” of the experts’ expertise
    4. Evidence of the expert’s interests and biases vis-a-vis the question at issue
    5. Evidence of the expert’s past track-records

See also: testimony