How should we distinguish good appeals to authority from bad ones?

Bad Appeals to Authority

• Appeals based purely on somebody’s profession, or social status. Police officer, priests, academics, politicians will lie or make false statements for all sorts of reasons. They aren’t a source of truth simply by virtue of wearing a uniform, having a position of authority or even having studied. A scientist isn’t a reliable authority because they are a scientist but because of the process the claims they are making have gone through.
• Appeals to authorities past their sell-by date. Sigmund Freud is a sound authority on the opinions of Sigmund Freud, Karl Marx is a sound authority on the opinions of Karl Marx but Freud is not a sound authority on modern psychology (the science has moved on and Freud’s theories have been largely debunked. Marx has fared even worse as it isn’t even the case that Marx is an authority on Marxism as the movement went off in a hundred different directions after Marx died. Charles Darwin is not an authority on the theory of evolution by natural selection – even though his work is still impressive and brilliant, the science has advanced far past what Darwin could have known.
• The decontextualized authority. Quotes, anecdotes, sayings from notable people stripped of their surrounding context including either the social/cultural situation in which they spoke or even the surrounding text.
• The fake authority. Both people who have been shown to be frauds and invented quotes or sayings of historical people fit here.
• The authority by virtue of being a great man. The gendered term there is intentional as it is likely to be a man. Thomas Jefferson often falls into the other points listed above, as a victim of made quotes, real decontextualized quotes, his views on 18th century politics poorly applied to 21st century realities, or simply an assumption of him being right by virtue of being a very notable US President. Additionally his being regarded as being a particularly sage and insightful thinker makes him somebody often cited.
• The talking our their hat authority. A noted expert may strongly assert an opinion perhaps on a matter of science – but if they are asserting that opinion just as their opinion without it being backed by research and peer review then it really doesn’t count for much more than ‘a notable person said something’. It doesn’t make what they said wrong and possibly it is something to pay attention to but it isn’t something you should put too much weight on.
• The not really saying anything authority. Petitions, list, signatories – again these kinds of things when signed by noted authorities aren’t entirely nonsensical or necessarily wrong but they don’t carry a lot of evidential value. They show a group of notable people agree with a position and not much more. They may indicate that you should pay attention to this issue but they aren’t themselves strong evidence.
• The authority by single experience. A single instance of a person from a given social or ethnic group is not necessarily an authority on the experience of the group in general. Of course they are an authority on their own experience.
• The degraded by transmission authority. Not every supposed fact that is cited is as sound as it looks. Sometimes zombie factoids live on independent of actual evidence. Such zombie facts then live on by being cited back and forth. Tracking claims back to their original source is important.

Good appeals to authority

• Referencing books by noted academics. A book can be wrong, an academic can be mistaken, lying, talking out of their hat or just plain crazy. However pointing somebody to a book allows for others to firstly check what the book says in context but also to find criticisms of the book and dissenting opinion. Evidence that can be further verified is good.
• Referencing peer reviewed scholarship. Peer review has its flaws and lots of peer reviewed papers have flaws or are published selectively (publication bias where negative or disappointing results don’t get papers written or published). However like the first point a peer reviewed paper allows others to check and consider the way a study was done or to find critics of what was done.
• Broad consensus of experts. Assuming this can be demonstrated and that the expertise is relevant, it is notable if many people in the same field all agree on a conclusion based on evidence and process.
• Witnesses to an event, individuals about their own experiences. Without evidence to the contrary (including the plausibility of the claim made) people usually tell the truth.