We all know that appeals to authority do not make for good arguments. At the same time, it seems that in daily life, people act as if they have actual knowledge and not belief. I was listening to a podcast on al Farabi, and it was mentioned for him, hearing something from an authority did not constitute knowledge. So that got me wondering:

Have any philosophers looked at how to reconcile how much in our daily experience we trust beliefs we get from knowledge and the philosophical intuition that those beliefs do not constitute knowledge proper?

4 Answers 4


I am going to provide an answer to the substance of this question but before I do I have to point out that it makes many common but false assumptions.

First the question assumes that some people or institutions are authorities whose statements should be trusted. This is false. Anybody can be mistaken about anything. There is no way of doing things that can guarantee any particular piece of knowledge is correct. another problem is that even if a particular piece of knowledge is correct, applying it to your specific problem can lead to errors. The value of experts cannot lie in them not making errors, nor in any practice such as trusting what they say, that makes no sense except in the absence of error.

There is another associated problem: how do you decide who is an expert? If you assume that somebody is an expert because the government, say, claims he is an expert then you have laid yourself open to being bilked. A government official might say somebody is an expert because the expert claims the staff in his department, department X, should be increased and he likes having lots of people around so he can socialise with them. Or the government official might have a subjective feeling that what he does is important when in fact it isn't and select experts who agree with him. Or experts who don't agree that department X does an important job might decline to work with the official in charge of X. Other institutions that are deemed to be authorities suffer from similar problems.

But this is just a particular example of a more general problem. People make a lot of choices using tradition: information about what is true and how to act created by people in the past. It is impossible to avoid doing this. You cannot recreate for yourself all of human knowledge.

The standard theory of rationality requires that you do precisely this. It demands justification: showing that knowledge is true or probably true. And just taking somebody else's word for it doesn't come anywhere near to being adequate by that standard. Nor does reading something in a scientific paper do anything at all to help. It's just a piece of paper. Anybody can write anything they like on a piece of paper. Worse, the standard of justifying positions or practices can't be met at all. Arguments always have premises and rules of inference that haven't been shown to be correct or probably correct or anything like that. So no argument can justify anything.

But there is a better standard of rationality that can be met: the critical rationalist standard. All position should be held open to non-justificational criticism. When assessing a position you should consider whether it solves the position it purports to solve, whether it is consistent, whether it makes sense in the light of other knowledge and that sort of thing. (See "Realism and the Aim of Science" by Popper, Chapter I.)

What would a critical rationalist theory of tradition look like? Karl Popper started to explain this in "Towards a Rational theory of Tradition": Chapter 4 of "Conjectures and Refutations". When you have a problem you should look for flaws in the traditions you are enacting that seem relevant. If you find a flaw you should then try to fix it by coming up with proposals for how to fix it and criticising those proposals until only one is left. In this way, you don't throw out stuff unless you have a criticism of it, but you are open to replacing anything that doesn't survive criticism. Tradition, including stuff that people uttered by people who are said to be experts, should be accepted if it seems unproblematic, but it should not be trusted.


If you haven't been introduced to the term before, you may want to look up epistemic authority. Many philosophers have dealt with this problem throughout history, in one guise or another. David Hume was arguably the clearest in holding that sense-data is the only validator of testimony by other people. In one way of reading him, our senses themselves are the final authority for coming into knowledge.

When we point out that an argument is fallacious for appealing to authority, what we normally mean is that a claim or premise has been presented as a fact, with no other basis than trust in the source. This trust alone does not provide sufficient reason to establish a claim as true: for anything which is the case, there is a reason why it is the case, and a citation cannot fill the same role that a reason would. This presents a problem when we trust our senses to be authoritative, in the same manner we see a problem in appealing to expert testimony to be authoritative. It's not always unreasonable to trust in an authority, but it's not sufficient to rely on an authority alone for establishing the truth of things.

  • I'm adding this as an additional note, which I came across today and I think contributes a bit to my answer, from an excellent article by Patrick Stokes: "It’s a common misconception that appeals to argument are automatically fallacious. They only become fallacious when a) the authority is irrelevant to the question and/or b) we use the appeal to claim more than it can do." -- From a reply by Stokes to a comment far down the page of his article here: theconversation.com/no-youre-not-entitled-to-your-opinion-9978
    – Ryder
    Commented Aug 9, 2014 at 11:45

Many appeals to authority are really appeals to reputation. The reputation of the stakeholders, any institution, and the author. Within the scientific community their are partisan camps and charlatans but relatively few due to peer review, scrutiny, and the scientific method. This is a principal difference between science and religion (or medicine and alternative medicine). If a party tries to assert authoritative knowledge without being able to back it up with evidence they will quickly be called out for it and their reputation will suffer. At the level of institutions or industries - how much trust does climate research by big oil or health research by the tobacco industry get? How much trust is given to national governments with a reputation for media manipulation or spin?


Let's all be philosophers who look "at how to reconcile how much in our daily experience we trust beliefs we get from knowledge and the philosophical intuition that those beliefs do not constitute knowledge proper."

Perhaps your question asked about an existing philosopher who has done this because your interest in philosophy is heavily tilted toward a reliance on authorities ;-).

When I heard that Socrates made the claim that we cannot know anything, I liked that claim. I have taken it up as my own claim as well. What we can do is define things, and if we provide definitions that are simple enough, then others can use the same thing in the same way and we can achieve some agreement. For example, we defined 2 and + and = and 4, so we can agree that 2+2=4, and we all "know" what it means. Well, we all define what it means. But we use the word know to mean something, and we would be well served to be more honest about it. That's what I think your question seeks: honesty.

I have three children. They and my wife used to ask me if I was sure when I answered a question, sometimes. At first (and probably long before I got married), I would put a sureness-measure on my answer: "99.44% sure." Since I liked math, I imagined I was saying that the present circumstances could present themselves ten thousand times, and I think I'd be right in all but 56 of them. I suppose that I have decided that I will claim knowledge when the chances I estimate of being wrong are one in a million, though ten thousand also seems big enough, to say I know it. I still say I'm sure when it's one out of a hundred or more, but by now everyone I talk to knows what "I'm sure" means when I say it.

Ok, so let's reconcile authority with knowledge. There are two kinds of authority, one of which is actually fake. The fake authorities say "It's my way or you go to jail." They use brutality and violence rather than reason. They suck. Can we say that here? Well, they do, and that's why. The other kind is people who study so much that they're aware of a lot more in a particular area than the rest of us. If they disagree with what we say about a particular thing, they're probably right because... they're an authority. The reason that's a fallacy is that probably isn't good enough. To be good enough, you have to go through all the work the authority did (and if they did it well, that shouldn't take too long).

Fallacies are dangerous, but most people learn that they've used a fallacy by suffering (not dying), so they aren't all that dangerous. The fallacy of appeal to authority is difficult to see (but it's getting easier because of fake authority) because most authorities (but see my previous parenthetical) actually do pretty good work. If they use threats instead of reason, then you can write them off. Knowledge does not come from coercion.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .