I have an idea as to why it might be valid, but I would like confirmation, rebuttals, or maybe just better answers.
One problem with "appeal to authority" and all informal fallacies is that they are problems of application. Thus, the fallacy is really illegitimate "appeal to authority" or we could say punting to authority what we should decide for ourselves.
The rub is that in arguments sometimes one party accepts an authority that the other does not, and then from the side of one interlocutor we have a fallacious appeal to authority and the other not.
E.g., An argument between a Catholic and a Protestant Inerrantist. For the Catholic, the pope is a legitimate authority and the bible without the magisterium's interpretive mechanism is not. (The Inerrantist believes the opposite).
Thus, we'd have cross-wise accusations of "appeal to authority". Or to put it another way we'd have the "name it as a fallacy fallacy" which commonly occurs with informal fallacies.
Presumably appealing to authority is non-fallacious precisely when it makes sense to believe the person offering the information is reasonably expected to be knowledgeable on the topic. In other words, it's a question of what for us would qualify as an "authority."
In this sense, it's going to tie into more basic epistemological considerations about knowledge. A general skeptic would have to say no one is "reasonably expected" to be knowledgeable about the matter at hand.
Throwing out skepticism, one criterion appears for technical skills: objective success in matters relating to what they are also commenting on. Thus, I'd for instance take a successful carpenter to be someone I'd trust on matters of wood-working, an effective doctor on matters of medicine, etc.
For theoretical things, it's going to be harder to know whether to accept the authority or not. A further criterion might be: are they accepted as an expert (in general? in my community?).
In arguments, we can adduce a further thing: do we both agree here and now to treat this person as an authority?
The validity of an appeal to authority depends on what forms of justification are accepted for a statement. In some extreme cases, the only acceptable justification may be a First Order Logic proof based off of agreed upon axioms. In such a case, an appeal to authority is not an acceptable justification. However, in many cases, we accept a justification which is less complete than this. We may be in a hurry, and not have the time to vet a logical argument ("When I say jump, I expect you to jump, and only pause to ask 'how high' once you're already in the air!"). Or perhaps we may have some minimal question about the validity of our axioms ("Are we 100% certain we'll go to heaven when we die?"). The rules of logic are notoriously fickle about dealing with axioms where there is any degree of uncertainty. For an excellent example from mathematics, one can look at the debate over the Axiom of Choice. The Axiom of Choice seems highly intuitive, but has some behaviors which are so extraordinarily counter intuitive that many abandon the axiom all together.
So what do you do in these circumstances where one cannot use some crisp clean logical proof, for one reason or another. Given that one cannot prove the claim, what other "good" solutions are there which don't depend upon a proof? One may consider the power of belief in such a search. It is trivial to come up with scores of examples where the belief that something is true leads to a desired outcome, even if that something was not actually true. Of course, it is also trivial to come up with scores of examples where the results of said belief end spectacularly poorly. However, it points to an alternative to being able to prove an argument. If one can defend the claim that believing in that argument is "good," potentially by arguing that the outcome of such a belief would be good, then one may win the argument. One wins the argument not by proving themselves wrong, but by convincing the other party that it is no longer in their interests to fight. The validity of such arguments is founded in their ability to cause "good" results through the belief in an idea, whether or not that idea was true. If you completely reject such arguments, you are limited to beliefs which are logically provable, and there are some serious complications to consider some time down the road.
An Appeal to authority is one such example. In many cases we come across in our lives, we simply do not have all the information needed to create a decisive logical argument. Worse, we've observed that it is very easy to create proofs involving false axioms to convince us, because we lack the fundamental background to weed the true axioms from the false ones. I can't count the number of times I've seen an anti-gun argument that starts from the assumption that taking guns from law abiding citizens will also take them away from criminals. I can't count the number of times I've seen a pro-gun argument that starts from the assumption that taking guns from law abiding citizens will have no effect at all on gun owning criminals. I do not have the background to tell which argument is right, and have a gut feeling that the answer is somewhere in the grey region between. However, I have a suspicion that others may know information I do not (this suspicion is reaffirmed on a daily basis... or an hourly basis on particularly bad days). Someone else may have the information I need.
The only question remaining is how do I identify the correct person's opinion to "believe to be true?" An Appeal to Authority is one approach to this. The idea is to identify a person who:
- Likely has a more comprehensive understanding of the problem than I do.
- Is not likely to maliciously misinform me for personal gain
As it turns out, society has a general tendency of putting such people into positions of authority. It's certainly not perfect, but one must admit the general trend. There are more than enough corrupt CEOs out there, but the vast majority of CEOs run businesses far better than I would if I were in their position! Thus, if I find someone in a position of authority on a topic, there is a reasonable chance that believing what they claim (regardless of its truth value) will yield a good result.
I will not claim the process is perfect, for it isn't. However, it does seem to have a remarkable history of success despite its notable failures (such as many genocides). One reason for its success is that it is not dependent on the truth value of the argument; rather it is dependent on the goodness of the belief in said argument. While the truth value of many arguments of interest is a static value, belief can come and go. Consider an ultimate extreme case in a wartime scenario. A lieutenant orders a sergeant to take his fireteam to a position. The sergeant issues orders to his team, and they act. In taking that position, one of the soldiers under the sergeant's command is killed, but the position is taken. Now consider the orders from the lieutenant to the sergeant. The sergeant does not have a logical proof explaining the outcome of such a move. In fact, if the sergeant took the time to question the lieutenant, that time may cost them the battle. War is a very dynamic environment, and there may only have been a window of a few moments to take that position -- a window lost if the sergeant questions orders. Accordingly, the sergeant relies entirely on the lieutenant's position of authority to decide how to act. He assumes that believing that the lieutenant knows what they're doing, and the decisive action which follows such a belief is "good." The same goes for the sergeant's orders to his fireteam. They must believe the sergeant know best, even if that is not actually true. A lack of belief would cause hesitation which is simply not forgiven in the horrible landscape known as war.
In the aftermath, there is time to reflect. The poor solider who gave his life for his country is obviously not reflecting much, but the sergeant certainly is. He issued an order to a man whom he knew was conditioned to believe such orders were good, and that order got the man killed. In the reflection, he now has time to explore the painful question of "did he have to die?" The sergeant has no proof that the lieutenant's orders were actually good ones, and now that the firefight has ended, the value of blindly trusting in the belief of those orders decreases. Now the sergeant may question the lieutenant, striving after the truth value of the question plaguing him: "Did my fellow man have to die?" He may not even like the answer he finds. However, the validity of the appeal to authority did not depend on the truth value of the orders. It depended on the momentary value of the belief in those orders.
One way to justify an appeal to authority is to define expertise as follows: a person is an expert on a topic T, if and only if, he has authoritative and comprehensive knowledge on the topic.
While such a person likely does not exist, as there's often something that's unknown even to those who are amazingly well versed on a topic, yet known by someone else, we can reasonably assume that those who have gone through extensive education, who have defended their positions repeatedly, etc are experts.
So why does appealing to expert opinion work? If we assume that the person is an expert, then he knows all that is known about the topic (comprehensive knowledge), understands the topic fully (authoritative knowledge), and is honest enough to not make a claim that is not known, by him to be true. Therefore, assuming expertise, we can safely utilize the claim made by the individual.
There is such a thing as the division of labour, and what goes for the labour of the hands goes for the labour of the mind: that is intellectual labour; one cannot, not in todays world, and probably not ever, be in a position to verify everything for oneself, hence appeals to authority are neccessary; the point is finding authoritative authorities: authorities that are valid.
In fact we can render the statement tautologous:
An appeal to a valid authority is always valid
removing the first 'valid' qualifier means that this sentence becomes:
An appeal to an authority is sometimes valid
Yes it can indeed be valid. We employ people all the time for them to exercises there authority in our lives.
When you go to school, your teachers who have gained a certain amount of authority trough years of study, exercise there authority to the end of teaching you enough Geography so you can pass your exams.
When you go to your local GP and ask him about this rash on your stomach, he exercises his authority in medicine to prescribe to you the necessary ointments, as to cure the bad case of eczema you have.
This is all reasonable and logical.
Where an appeal to authority becomes fallacious is when an argument is proposed whereby the claim is said to be more true solely because it agrees with some authority. Authority can be wrong after all.
An appeal to authority is almost never valid. However, if it is declared as such and accepted by both parties, it is alright to use it. First thing to know is that it is an informal fallacy. So it is wrong or right depending on the way it is used or more precisely what inference does the person using it wants the other party to draw.
If the inference to be drawn is about the truth of a belief or proposition, then such appeal is never valid. If the person putting appeal to authority uses it to justify his belief or having a particular belief, then it may be not be fallacious to use it. Here hidden assumption is that it is justified to have those beliefs that are held by experts in a field.
An appeal to authority is not a valid A PRIORI argument. The only way an appeal to authority would be even useful is if you are unable to evaluate the argument itself (that is, it is either hidden or you are unable to make valid inferences about arguments!). The appeal to authority is an a posteriori inductive inference.
An appeal to authority is never valid under any circumstances. An idea is either true or false. What is relevant to the idea is whether there are known criticisms, such as inconsistencies or clashes with experimental data. There is no privileged source of knowledge that always gives gives correct answers, or probably gives correct answers or anything like that. We can easily make mistakes when measuring something. We can easily make mistakes when interpreting something a person wrote. We can easily make mistakes when doing a measurement or making an argument. As such, every source of knowledge is fallible. Nor is there any way of separating right wrong arguments other than critical discussion. All sources of knowledge should be treated as fallible: no exceptions.
See "On the sources of knowledge and of ignorance" by Karl Popper in "Conjectures and Refutations".
The only rational way of dealing with a source is to explain its significance and to be prepared to defend its arguments. Citing a source as if it ends an argument is anti-rational. For more substantive guidance see
To put this another way. Authority is a social status, which has nothing to do with being right or wrong, and so is irrelevant to assessing a position.