Arguing with Trump supporters: associative thinking

Many years ago an acquaintance told me that she was voting for Reagan because his secret security agents loved him. I told her that Hitler’s security guard loved him. She was offended that I had said that Reagan was Hitler.

I was trying to make a point about her major premise. That is, I was looking at her argument from the perspective of argumentation. I was trying to say that her major premise (if your personal guard likes you, then you are a good person) was not one she believed. She understood me to be making an associational argument (which is how she thinks).

For both of us, the question was, how do I know I’m right? For both of us, the question was, can I find evidence to support my belief? For me, the question was, do I agree with my own major premises? For her, the question was, is this thing I believe to be good associated with other things I believe to be good?

Far more people approach issues the way she does than reason the way I do. And, in fact, her way of thinking makes sense under some circumstances. I like this store, and they have this product, and it’s a lot like another product of theirs I like, so I’ll probably like that product too. This person recommended this movie, and I’ve liked other recommendations they’ve made, so I’ll probably like this movie too.

Liking this kind of store, or this kind of movie, or this kind of restaurant, and therefore wanting to find stores, movies, or restaurants like that is a good enough way to make some initial guesses. What I mean by “good enough” is akin to what Winnicott said about “good enough” parenting. If we set our standard as perfectly rational reasoning at all times, we’ll never make any decisions. Relying on associations is a good enough way to start. It becomes problematic, and actively damaging, when

1) our chains of association are unreasonable. For instance, I might decide that, since you and I both like a restaurant (“Chester’ Burgers”), then I’ll like a different restaurant you recommend (“Hubert’s Vegan Noodle House”). That’s a bit weak—it would probably be a stronger way to think about the situation if it’s unusual to like that first restaurant, or you like all the same restaurants I do, but it’s good enough to give the second restaurant a try. If you and I both like a movie, or we vote the same way, or I think you’re a “good” person, then the relevance of your restaurant recommendation is much more tenuous.[1] A lot of people assess the world in that tenuous way (I’ll come back to this)—they decide that people are good or bad, and that a good person is good in every way, and every good thing is associated with good people. Therefore, a person who is a good actor must also be a good President.

2) the preliminary conclusion we come to is then protected at all costs (we aren’t open to changing our minds). The second flawed application of this way of reasoning is if I try the restaurant, and have a bad experience, but refuse to admit it was a bad experience (or decide that you didn’t really like the first restaurant).

3) we rely more on the negative associations than the positive–for instance, if I refuse to go to a restaurant you like because there is a restaurant you like and I don’t. This one also has lots of exceptions, depending on how much alike the two restaurants are. It makes less sense if I won’t go to a restaurant because you like a movie I don’t like, or vote differently, or I think you’re a “bad” person (with the reverse of some of the exceptions noted about the first method).

4) we use this method for all situations (this is particularly damaging if it’s connected with the first). The fourth is self-explanatory—there are times that this kind of associative thinking isn’t very useful. A person might be a very nice person, but that doesn’t necessarily mean that they’re giving you good medical advice.

Obviously, a tremendous number of decisions are based on these sorts of associations. The best way to think about this sort of decision-making is that associations are most useful when they are treated as analogies. The more that the next case is like the previous, the more relevant your judgment is. Thinking about these associations as analogies means that we think about them in terms of degree—how much is this restaurant like the other one? The more that I understand your reasons for liking the first restaurant, the better able I am to determine how much your judgment applies to other restaurants, let alone movies or Presidents.

A lot of people who think associatively don’t approach the associations as analogies, but as equations. This method of thinking is what Chaim Perelman and Lucie Olbrecths-Tyteca called “philosophical paired terms.” Very few scholars of rhetoric pay attention to them because thinking via paired terms doesn’t fit our normal schemes of argumentation, but they’re important to understand because, if someone is “reasoning” this way, then normal schemes of argumentation won’t work. It took me years to figure that out.

And Hitler analogies are a great way to explain the problem I kept running into when trying to argue with someone who thinks via paired terms.

Normal schemes of argumentation are (more or less) relations of causality. So, the claim that “Reagan is a good President because his secret service agents really like him,” to me, was a claim that there is something that causes secret service agents to like someone that also causes that person to be a good leader. Hitler’s security liked him, and he was not a good leader, so the Hitler example shows there is a not a causal relationship.

Once I realized that people weren’t thinking in causal terms, I thought they were reasoning by signs. In other words, I thought they believed that having a security team that likes you is a sign that you are good. And so I thought we were talking about whether something was a perfect or imperfect sign. I tried to show that it was an unreliable sign, but that didn’t work. And it didn’t work because the person couldn’t get past that I had said Reagan was Hitler. And, of course, I’d said no such thing. People like that really seemed to have trouble understanding the concept of analogy, especially the concept mentioned above—that analogies are valid to varying degrees, depending on the degree of likeness. If you are saying this person is like that person, you aren’t saying they’re the same.

For people who think in paired terms, chains of association are links that equate the two linked identities, concepts, people.[2] They divide the world into good and bad, with all the good things chained to each other and in opposition to the bad (which are also chained to each other). Bad people must be bad in every way, and associated with every other bad thing, and good people must be good in every way, and associated with every other good thing.

Thus, as with another way of reasoning (discussed elsewhere), it’s possible simply to attribute a quality through opposition. What I mean by that is that if I believe that everything can be divided this way, and I’m Christian, and you disagree with me (about anything), then I conclude you must not be Christian. It’s this way of thinking that causes some Trump supporters to accuse everyone who disagrees with them of being a socialist.

There are two ways to try to persuade someone who is thinking in terms of paired terms. Perelman and Olbrechts-Tyteca describe one method—dissociation. You can see MLK engage in dissociation in “Letter from Birmingham Jail,” in which he tries to take the paired terms of the “Eight Alabama Clergymen” and dissociate them. Loosely, their argument associates their policy (cease protests and try to work slowly with the new Mayor and City Council) with: patience, peace, lawfulness, faithfulness, “insiderness.” Each of those terms is contrasted to the ones they try to associate with King (an outsider, they insist): impatience, violence/conflict, lawbreaking, communism (this one implicitly). King slowly and patiently (thereby performatively refuting one of their claims) argues that his organization’s position shows real patience, peacefulness, and so on, while their position settles for apparent (superficial) instances of those values.

This method generally leaves in place the binaries, and tries to persuade the person they’ve miscategorized things. It didn’t persuade the Eight Alabama Clergymen, but it did persuade others.

I think the reason it didn’t persuade them is that the most important (albeit unstated) term associated in those paired terms is: me. Reagan is good, and I am good for liking Reagan (and for recognizing his goodness). For people who live in the world of paired terms, if you aren’t good, you’re bad. So, by saying Reagan wasn’t in the good category, I was saying she didn’t have good judgment, and she isn’t good.

Paired terms provides a kind of clarity and certainty that is breath-taking. So, the second way of trying to argue with someone who believes that Biden must be a socialist because he isn’t a Republican (an instance of thinking via paired terms) is to try to persuade them that that way of thinking is flawed. As you can guess, that isn’t easy.

You’re trying to take away their clarity and certainty. And you’re telling them that they’re not only mistaken in this instance, but in their whole view of the world. You aren’t just saying that Reagan might not be good; you’re saying they might not be good.

In my experience (I’ve never seen any studies on this issue), people who think this way are often suckers who’ve spent (lost) a lot of money on dodgy cures, bad investments, and get-rich schemes, over and over. I think it’s because they can’t learn from their mistakes, since they can’t admit that this way of making decisions isn’t working for them. They’re often Followers, who idolize (and believe they have perfect insight into) various celebrities. This is also the sort of person with whom I have made the least headway in terms of persuasion. Consistently.

The only thing I’ve sometimes managed to do is point out that connections between two terms are invalid, such as showing that people who claim to be Christian are violating basic dicta of Christianity. But that usually causes them to block me, or to say, “Democrats are the ones who kill kids.” And we’re back to the pretense that Republicans care about abortion.




[1] I can imagine situations in which these connections wouldn’t be so tenuous—perhaps we like the movie because we both really like the star, and you mention that star often goes to the restaurant, or the owners of the restaurant contribute a certain amount of profits to the political party we like, or I think you’re good because you only eat at restaurants with ethically-sourced food. But they’re pretty unusual, if not actually rare.

[2] Another way to think about this is in terms of Venn diagrams. I would map the diagrams as “people who are good leaders” intersecting with “people whose security detail likes them,” but for some people there are only two circles, and they don’t intersect at all. There is the circle of good things/people/concepts, and the circle of bad ones. Having your security detail like you puts you in the good circle. If it puts you in the bad circle, as she understood me to have suggested, then I must be putting Reagan into the same category as Hitler, meaning I’m saying that Reagan killed six million Jews, and he didn’t.