Authority? What authority?

Look, I know what I’m talking about.

Have you ever said that? And has anyone ever said that to you? It’s an appeal to authority, and, according to some people, it’s an instant fail: the argumentum ab auctoritate (argument from authority) – a famous logical fallacy!

Except when it’s not. Because if appeals to authority were always fallacious, our entire legal and educational systems would be voided. Among many other things. 

I’ll explain.

In its simplest form, the argument from authority puts eminence over evidence: “I am an authority on this subject; therefore, anything I say is true.” This is easily seen as false; nobody’s right all the time, and knowledge advances by proving experts wrong. And while anyone who’s learned enough to be an authority should know this, sometimes they just don’t want to have to deal with pipsqueak upstarts. In this “shut up, kid” case, we could equally call it the “because I’m the mom” argument (argumentum ab matre: “argument from mother”) – an argument as much from status as from knowledge.

And it is annoying being on the receiving end of it. Of course it is. I remember when I was in first year university and the teaching assistant for a stage production course mentioned that, when using a reel-to-reel tape recorder (hey, it was 1984), you could change how long a piece of music played by adjusting the tape speed by a few percent. I pointed out that that would put it out of tune with any live instruments (we had just been talking about having taped plus live together). She said no, it would not. Now, I knew perfectly well that it would, not only from the principle of how analog audio tapes work but from having worked with reel-to-reels because my dad did audio-visual production. But when I objected, she just shut me down: she was the TA, a graduate student, and “I think I know more about this than you do!”

That’s obviously an improper use of argument from authority, but you will notice that it really was a simple argumentum ab matre, drawing entirely on her status. I had relevant knowledge and experience, but I was prevented from presenting it. (I consoled myself with the knowledge that she would likely find out her mistake soon enough.)

There are prominent instances of eminence prevailing over evidence, such as the case of Theophilus Painter, a zoologist of note who declared that humans had 24 pairs of chromosomes, and although there was increasing evidence that the true number was 23, for a long time Painter’s number was held to because he was the expert. This, too, is annoying, and it impedes the progress of knowledge. Getting trumped by a superior is (alas) common in academia, and any young researcher will likewise be vexed at an improper appeal to authority – even if in the end the authority is not wrong.

But now consider a kind of instance quite familiar to some of us, one not in the empyrean of academe but in a more casual interpersonal context, such as Twitter. Let’s say that someone (we’ll call him Logic Bro) says “You shouldn’t use the possessive, like ‘my students’ or ‘my friends’, because you don’t actually possess them,” and a linguist (we’ll call her Lingo Doc) says “That’s not how it works. The ‘possessive’ doesn’t actually exclusively, or even mainly, signify possession.” Logic Bro is not convinced, and it goes back and forth a few times until Lingo Doc says, “Look, I’ve tried to explain this; you’re making some very basic errors here. This is stuff I teach about in intro classes. For heaven’s sake, I have a PhD in this subject.” And then Logic Bro says, “You lose. You just appealed to authority. You didn’t even know that’s a fallacy? You don’t know much, do you? Ohhh nooooes are teh lingwists going to blockkk meee”

I think most of us would agree that this is annoying, but in the other direction from the previous examples. Is Lingo Doc wrong in appealing to her authority? If you say yes, I have some bad news about your diplomas and degrees, which are all attestations to learned and earned levels of authority. But let me approach this from two angles: the pragmatics of the interaction and the value of authority in discussing matters of reality.

First, the pragmatics. Whenever people communicate with each other, they are not simply making logical statements about the world. All communication is behaviour, it’s always done to have an effect on the person(s) you’re communicating with, and it always draws on and feeds back to a definition of the relationship between the participants in the interaction. Conflict situations very often involve competing definitions of that relationship. And that’s especially true in many (probably most) cases where an appeal to authority is introduced. 

When I challenged the TA, I was making a knowledge claim, which also happens to be a status claim, especially in a classroom context. It was important to the TA to maintain her status and not to have this mouthy kid putting himself above her. Likewise, when Logic Bro contradicts Lingo Doc, he is making a claim of greater intellectual status, because he is asserting that he is right and she is wrong – in fact, pretty much every single Logic Bro on the planet is mainly in it to assert dominance. He uses “logic” as a weapon, and the moment she does something he can characterize as fatal error, he treats it as a trump card and adds further insults. She naturally bristles at his conduct, not only because he’s rude and wrong but because he’s wrong and rude, and she knows and cares about the subject matter. So she asserts her authority in part to counter that, but also because she can’t teach an entire undergraduate course in linguistics to Logic Bro in a few tweets.

Which leads us to the value of authority. Authority matters because we don’t learn through simple syllogisms with the premises all laid out. It’s much messier than that. In real life, whether in classrooms or anywhere else, when we learn that something is true, most of the time we are really learning that it is most likely true, and we accept it as true until such time as our understanding is revised. And revision happens a lot in the course of education. When you study linguistics (especially at the undergraduate level), half of each course involves unlearning stuff from the previous course that was too simplified to be really true. Many other subjects are like that too. And as we learn about things, to increase the likelihood that what we learn is true we need to be guided by those who know much more about the subject, because we don’t even know enough to know for sure how we should learn about the next stuff, and also we can’t reinvent the wheel all the time. This doesn’t mean that those with more authority are always right, but the odds are greater that they’re right than that we are, and until we have a comparable basis of knowledge on a given topic, we’re not in a very good position to assess the truth of what they’re saying. If I don’t know a hammer from a screwdriver, it’s a waste of time to debate the finer points of carpentry with me.

We know this. Our whole society operates on the value of authority – not just our education system but our courts too: eyewitnesses have more authority on the basis of direct experience of the subject. So do our moms, for that matter. This is also what we’re appealing to when we say things like “Look, I know Jeff, and he wouldn’t say that.” We’re not always right, but we’re certainly not always wrong! Of course experience is fallible and subject to biases – many people who have been cooking for years mistakenly believe it matters whether the shiny side of the aluminum foil is up, for example – but lack of experience is generally even more fallible and subject to biases. When someone says “You never know,” more often than not they never know because they doubt rather than learn, while those they doubt do know pretty well.

So when Logic Bro comes in with an elementary failure of understanding and demands proof of why he’s not right in his assertion, he’s really demanding that Lingo Doc somehow impart to him a comparable basis of understanding of the topic in a very short time, for free, while he insults her at his leisure. It’s like trying to argue physics with a five-year-old. A debate on a topic has to take all the relevant knowledge on the subject as understood except what is being directly addressed, and someone coming into a debate without that basis may as well be coming into a casino and laying down a stack of money that is really all blank paper with just a single bill on top. Citing your experience and authority is like riffling through your own stack of bills and asking Logic Bro to do similar.

Let me put it one more way: If you agree with Bayesian inference, you can’t be opposed to the value of authority. 

Bayesian inference is an important thing in probability; it allows you to adjust your assessment of probability on the basis of prior knowledge of related factors. Let’s say, for example, you have a coin and you flip it 19 times and it comes up heads 19 times in a row. Many of casinos’ most profitable customers (biggest losers) would bet on its coming up tails the 20th time because it’s way overdue for tails. Logic Bro would tell you that the odds of its coming up tails are exactly 50% because in a random toss with a fair coin it doesn’t matter what the previous tosses are. A Bayesian kind of thinker, drawing on real-world knowledge of coins, would note that the odds of a fair coin coming up heads 19 times in a row are less than one in half a million, whereas the odds of the coin not being a fair coin are probably much higher. So the Bayesian thinker would ask to see whether the coin actually has heads on both sides, or whether it’s weighted, and, if unable to get that information but having to make a wager, would (barring further knowledge, such as that you might be a skilled coin-flip cheat) bet on heads.

Trusting the word of someone with a lot of education in a subject over the word of someone with rather less education in the subject is like betting on heads. Logic Bro may insist that he just wants the facts, and that facts and logic stand on their own – just as a single coin flip is not influenced by prior flips – but since he most of all wants to “win,” he’s (ironically) overlooking important prior conditions. Prior conditions greatly increase the likelihood that a person with a PhD knows better about something than a person with no formal education in the topic. Lingo Doc has spent years learning about linguistics; of course that ups the odds of her knowing what she’s talking about. It doesn’t mean that she’s going to be right 100% of the time, but if you say she’s wrong and she says you’re making a basic error in your premises, you should consider pausing to double-check your understanding of the subject – because in contradicting her, you’re asserting your own authority, so you really ought to know what you’re talking about.