Why can't people agree?

It's mostly because people suck, but there's some altruism in there as well

According to Aumann’s Agreement Theorem people should never ‘agree to disagree’, they should always be able to come to consensus. So why don’t they? There are several big reasons for this, some having to do with people sucking and some with them being good.

The first reason really sucks: People lie. Often people want to win an argument either because they have some ulterior motive or because they just like winning arguments. Against a completely credulous opponent they can do this by lying or bullshitting about the strength of the evidence in their favor. The subtle but important distinction between a lie and bullshit is that in a lie the person saying it knows it to be untrue while with bullshit the person saying it doesn’t know if it’s true and often doesn’t care. With the most extreme bullshit reasonable bayesian priors would suggest that it’s overwhelmingly likely to be untrue, and hence probably should qualify as a lie, but people who are bullshitters often don’t understand that concept. Maybe lying and bullshit aren’t that different.

Ulterior motives are easy to come by. For example, one can say that if you give me money good things will magically happen to you. It’s easy to gain overwhelming evidence for this via simple selection bias, remembering people who it worked out for and forgetting ones who it didn’t. You may laugh at this as being rhetorical, but prosperity gospel is a whole industry.

(I’m probably showing a form of bias here myself by assuming that other people are minimally competent at data science fraud. If you want to commit fraud the hard-to-detect way of doing it is to throw out some of the data you’ve collected, that way all the data you present looks legit because it’s literally real data. Instead even Harvard professors do their fraud by the data science equivalent of scribbling on a map with a sharpie.)

The second reason for a breakdown of agreement has to do with who’s an authority. One common misidentification of an authority figure is the first person singular, that is to say, people think their opinions deserve the benefit of the doubt and the onus is on others to provide evidence debunking them. They will even accuse others of engaging in a ‘logical fallacy’ and being ‘dismissive’ for, well, being dismissive. They are of course the ones engaging in a logical fallacy themselves. People bullshit enough that bayesian priors should suggest that anything strange someone suggests without justification is probably bullshit.

It’s important to point out that appeals to authority are not a logical fallacy. For most matters of science, engineering, politics, and history only a tiny fraction of the general population have much hope of contributing meaningfully to them and everyone else is far better off identifying who appropriate authorities on a subject are and deferring to them. You might notice that the people I’m arguing with here like calling things logical fallacies. One even might say that these overly detailed breakdowns I like doing are mostly explaining things most people find obvious for the benefit of autistics and are only really useful for the very small number of people who simultaneously struggle to understand the world, but very good at thinking about things in the abstract and are earnestly interested in explanations. There might be a lot to that.

The other big common misidentification of authority is in the form of deciding people to entrust based on prejudice and superstition. Deep in the human mind there’s programming to be part of a tribe and view the tribal leader as the great authority on all things. The people designated by these means tend to be (but are no means exclusively) tall, charismatic, old, male, and have a penchant for righteous ranting. Religion obviously exploits this, but so do the mostly fake management consulting and CEO industries. Who should be viewed as trustworthy is a fraught and difficult subject, but you can do better than most people by using reasoning based on evidence instead of your gut instincts, particularly looking out for the biases mentioned above. Yes I know righteous ranting is fun, sorry to be a downer. There are whole industries based off exploiting your poor instincts, and while your primal judgement is good at many things, picking out trustworthy people isn’t one of them.

On a more positive note, some disagreements are due to subject matter experts sacrificing their own accuracy for the greater good. Received wisdom needs to come from somewhere, and prior to its existence experts can individually give their opinions but can only come to general consensus by comparing with other experts, arguing their positions, and weighing the opinions of others. Even after conventional wisdom is achieved it still needs to be questioned and reevaluated, especially as new evidence comes to light. This results in experts actually being more willing to break with the conventional wisdom, both because they’re more likely to have coherent reason to do so but also because they’re engaged in a general process of making the conventional wisdom more accurate. Interestingly experts are often unaware of the reasons for the convention of not simply giving the conventional wisdom, and out of arrogance or habit will explain current internal debates of their field to outsiders instead of switching modes to giving the conventional wisdom. This gives outsiders a lot of confusing and generally irrelevant information and gives them the impression that there’s a lot more internal strife in serious fields than there actually is. If you are an expert and are asked to explain something to the general public you should give a trailing view of the conventional wisdom and maybe hint at current discussions but say they’re speculative.

Please note that watching a lot of youtube videos on a subject does not make you an expert. Also note that disagreeing with the conventional wisdom is not evidence of your expertise. Very few people are experts on any given thing, and everyone is an expert on only a few things. Expertise at the level of it being reasonable to challenge the conventional wisdom is a very high bar.

Finally people sometimes disagree with the conventional wisdom because of the emperor’s new clothes phenomenon. At some point the conventional wisdom is so obviously wrong that you should question it even if you aren’t a subject matter expert. This is a very high bar, but does sometimes happen. The big example from my experience was reading Freud, who I vaguely accepted as the big authority figure in psychology even though his theories sounded slightly loopy until I read his actual writings and realized they’re utterly deranged. Unsurprisingly Freud has been thoroughly discredited since. Thankfully with ever greater communications technology instances of whole fields being out to lunch are decreasing in frequency. More common now is there being multiple ‘schools’ of a field, where at least one of them has some real credibility. Unfortunately it isn’t always the largest school.