There are an incredible range of subjects that people disagree about, but only a small number of core reasons that people disagree. When we encounter complex and difficult to resolve disputes, it can be helpful to break them down in terms of these reasons. This process can help give us insight into what is preventing a consensus from being reached.
Disagreements can be caused by:
1. Facts. People have access to different information. One person has studied physics, another hasn’t. One has spent years reading American news sources, another, Chinese ones. Differing sources will expose people to different facts about the world, which can naturally lead to disagreement. People have had different personal experiences as well, which can lead them to have different world views.
2. Definitions. Words are used differently by different people. One says “God” to mean the god of the old testament, another uses it to mean a guiding force in the universe. One says “art” to mean what is considered valuable by the art establishment, another to refer to creative works that evoke emotion. Typically, people do not even know precisely what they themselves mean when they use words. This isn’t a problem most of the time, but can cause confusion in certain cases. Suppose I take a chair and table, cut them both in half, glue one half of the table to one half of the chair with the cut edges touching, and paint the whole thing gold. Is the object I’ve constructed a table, a chair, both or neither? Our usual usage of the words “chair” and “table” are not well specified in this case, so we have trouble answering the question. I don’t know precisely what you mean by “chair”, and I don’t even know precisely what I mean by it. When I hear the word it triggers associations, memories, an image, but not a precise definition. When people are using different definitions they frequently get into unproductive arguments about those definitions without realizing that they are merely using words differently. “That’s not art, it’s just an old urinal that someone found and put in a gallery.” “Of course it’s art, it’s hanging in a museum, and clearly evokes emotion in the viewer!”
3. Values. People value and care about different things. One person thinks that more life in the universe is inherently a good thing, and so is in favor of increasing the size of the population. Another would prefer less life to be created if that would keep the existing people happier. One man places value on being remembered after he is dead, while another person doesn’t value it because she knows she won’t be able to experience that state of the world. Utilitarians place value on the world containing less suffering, Kantians value a world where certain universal rules are not violated, many people place value on the survival of animals species (even in cases where their survival does not benefit human beings), and so forth. There is nothing strange about different minds producing different labels for what is valuable and what lacks value. It does occasionally happen though that a person’s values are logically inconsistent (for instance, the person claims to value X while disvaluing Y, without realizing that X and Y amount to the same thing). [Comment: Related to values are the issues of tastes, preferences and interests. If you have a taste for chocolate rather than mushrooms, then you have an interest in going to the chocolate store rather than the mushroom store. A disagreement might then ensue regarding which store you “should” go to between you and someone with the opposite preferences. When you say “should” here it may be merely expressing your preference, but the conversation can take the form of a disagreement over what to do.]
4. Signaling. Sometimes people disagree (or claim to disagree) merely to signal information to others. For instance, they might say they disagree with someone they don’t like to demonstrate their dislike, or they might agree with someone they like to make that person like them more. They might say that Shakespeare wrote fantastic plays not because they enjoy reading those plays, but because they want to be respected for being a literary intellectual. They might say that they love the Yankees because that makes them more accepted by their group of friends.
5. Failures of logic. If the people on one or both sides of argument are engaging in logical fallacies or other failures of reasoning, they may end up disagreeing as a result. One person may argue that individual humans will never be able to live for more than 200 years because it has never happened before, while another will claim that we will definitely one day have lifespans longer than 200 years having extrapolated from the increase in longevity over the past 10,000 years. Since both sides have fallacious reasoning in this case, there is no reason to expect them to agree (unless their false logic just happens to lead them in the same direction). We can also include in this category the case where one person doesn’t have sufficient reasoning ability to understand an argument, and so is unable to properly evaluate it. Then, that person may disagree with the conclusion of the argument as a consequence of the argument being inaccessible to them.
6. Information processing methods. If people use different methods for processing information, they may end up disagreeing, even if the methods that each person relies on is quite reasonable. Brains do not perfectly update the strength of beliefs when encountering new information. Given limited computational resources, and the constraint that our brains evolved via natural selection, approximations like imperfect updating of credence are to be expected. If two brains both have pretty good but imperfect methods of updating beliefs based on evidence, they may end up disagreeing, even though they are both about as reliable as one another. When we reason explicitly information processing disagreements can also arise. For instance, if we want to estimate the chance that a friend will cancel on us for dinner, it is not obvious how we should weight our knowledge about how often friends cancel on us in general, compared to our knowledge about how often this particular friend has canceled on us. We have more data on all of our friends, which makes that estimate more reliable, but the data on this particular friend is more relevant to the particular case. Ideally, all of the known information should be combined in some way, but one can reasonably disagree about the best way to do this.
7. Default beliefs. If you were to design an intelligent being, it would start off with certain beliefs (implicitly, if not explicitly) and it would be silly, if not impossible, to have it start by assigning a precisely equal probability to every possible state of the world. If the prior/default knowledge of two intelligent beings differs, they may come to different conclusions, even when they processed information in precisely the same way and have access to all the same information. If a human baby, an alien baby, and a robot were raised in precisely the same environment, they might end up believing substantively different things about the world merely because their brains have different built in assumptions. Their brains might, for instance, start out assuming different things about the nature of other intelligent beings.
Suppose that two intelligences (people, aliens, A.I.s, etc.) exchange information about their beliefs, and their reasons for having those beliefs. If they start with the same default beliefs, employ the same information processing methods, do not make errors when they explicitly reason, are able to follow each other’s reasoning, share the same values, use the same definitions, and have access to the same facts (for instance due to exchanging factual information with each other), then you should expect those two intelligence to reach agreement.
Note, however, that these seven categories of disagreement are not entirely independent from each other. For instance, the distinction between bad information processing methods and illogical argumentation can blur. A poor choice of definitions can also be a form of illogical thinking when those definitions are meaningless (“Let’s define a squircle to be a square circle”). And the distinction between default beliefs and built in information processing methods can be fuzzy as well, since the effectiveness of information processing methods may depend on assumptions those methods make about the environment. It is also a possibility that there are disagreements which arise that aren’t due to any of the seven reasons listed above (should any other types of disagreements occur to you, please post them in the comments below!).
Let’s see how this categorization system can help us understand real world debates. As our example we’ll use the debate about when abortion should be legal (a convenient choice because this debate has many facets and great complexity), taking into account the arguments frequently used by liberals and conservatives in the United States. A 2006 survey found that 60% of self-described “liberal democrats” thought that “abortion should be generally available”, versus 17% of self-described “conservative republicans”. Furthermore, 68% of these conservative republicans thought it should be either “illegal with few exceptions” or “never permitted” versus 23% of the liberal democrats. We’ll stick to considering this debate in terms of the first five reasons listed above that cause people disagree, since these are the five that are the most useful to consider in practice.
1. Facts. Liberals and conservatives tend to be exposed to different information. For instance, it would not be surprising if liberals were more likely to hear about ways in which women are negatively impacted by not having access to abortions, while Conservatives were more likely to hear about women who regret having had abortions. Most of us tend to spend time with people that share many of our beliefs, and to read news from sources that are at least somewhat aligned with those beliefs. That means that the facts and ideas we hear tend to support what we already think is true. We should expect the friends and news sources of liberals to propagate more information supporting liberal values, and likewise for the friends and news sources of conservatives. Even google reinforces this tendency, when we happen to structure our queries in such a way that the results confirm our beliefs, and when Google automatically customizes our search results to be drawn from sources that are similar to those that we visit frequently.
The obvious solution to disagreements over facts is to exchange information with those you disagree with. This works well when people on the other side of the debate trust your sources as much as you do, and when they are willing to allow facts to change their minds. In practice, both of these assumptions are frequently violated. People commonly disregard information that contradicts their own beliefs, and latch onto the information that supports what they already think. When you encounter someone doing this, you can try asking the other person “what information WOULD make you change your mind about this issue, if that information were actually true?” That at least has a chance of getting the person to pre-commit to changing his or her mind if the relevant information can be found.
2. Definitions. The abortion debate is frequently phrased in terms of whether a fetus can be said to be a “person” (or, similarly, a “human being”), with the argument going “It is wrong to kill a person, so if a fetus is a person, killing it is wrong.” This would seem to imply that if we can just decide whether a fetus is a person then the question will be answered. But this begs the question of what we mean by “person.” We all agree with that Barack Obama is a person, but the usual definition we use for person isn’t clear cut enough to decide whether that word applies to the fetus case. And even if it were clear cut, it wouldn’t matter. The underlying question is not whether a fetus is a person, but whether we place disvalue on the death of fetuses, and if so, how much, and why. Phrasing the question in terms of whether a fetus has the status of person is bound to lead to disagreement, since intuitions about what “person” means differ quite substantially, and when one person says “a fetus is a person” and another says “a fetus is not a person” they may well not being using the word person to mean the same thing.
A handy trick for helping to resolve disputes over definitions is to taboo your words. Basically, this means replacing any confusing or fuzzy words with other words that get more closely at what you are trying to say. So, rather than “person” say “being that can suffer”, or “creature that cares about whether it lives or dies”, or “thing that is or one day will be an adult human”, or whatever else it is that you are really trying to denote with that word. In other words, if you’re stuck debating whether something is or is not a person, ban the word “person” from the conversation and use a more descriptive phrase. Now see if that allows the conversation to move forward.
3. Values. Some people place negative value on suffering, and believe that aborting a fetus is wrong if it causes the fetus to suffer, and that it is not wrong if there is no suffering. Others place negative value on the death of conscious entities, and so oppose abortion at the point that a fetus can reasonably be said to be conscious. Still others oppose depriving a being of its potential for future experiences, and so are anti-abortion on these grounds. Some people value rights, and believe the mother’s right to look out for her own interests trump those of the fetus until the point when the baby is self aware enough to get rights of its own. Still others value the life of anything that will become a human being, which includes fetuses.
Even though people can value different things without either of them being wrong, it still is sometimes possible to change what people are willing to say they value. For instance, you can try to demonstrate to them that one of their values conflicts with another of their values, which can cause one of those values to shift. Or you can try to show them that they do in fact value something that they claim not to, and prove it by means of an extreme example designed to provoke emotion.
4. Signaling. Abortion debates sometimes devolve into claims that those in favor of abortions are “against life” or that those opposed are necessarily “against women”. Obviously those who are pro abortion almost never actually want fetuses to die, they just want the mother to have a choice of whether to give birth. Likewise, there are of course reasons one may oppose abortion due to concerns about the loss of life of the fetus while simultaneously supporting rights for women more generally. When we use phrases like “against life” or “against women” they typically serve to demonstrate our group identity, and label the other side as bad guys. They say a lot about our emotional reactions and how we want our side and the other side to be viewed, but little about what the other side actually believes.
5. Failures of logic. Bad arguments are a prominent part of the abortion debates.
Some such arguments fail because they rely on the ambiguity of words and assumptions that are left unexplained. For instance: “fetuses are human life, and human life is sacred, so killing fetuses is wrong.” Some take logical leaps that are not sufficiently filled in: “since an early term fetus can’t survive without the mother’s body, it is the mother’s choice whether to kill it”. Other arguments are true, but don’t demonstrate the claim they are used to support: “abortion is dangerous to mothers because it caries health risks”. Of course, this leaves out the fact that child birth itself carries substantial risks to the mother. Other arguments lead to question begging: “A women has a right to control what happens to her body, so she is justified in having abortions.” Where does this right come from, and if there is such a right, why does it trump other rights that might come into play in this case?
Convincing other people that their arguments are illogical, and identifying failed logic in our own arguments can be quite hard. A helpful tactic for dealing with illogic in others is to ask them to clarify their argument, rather than outright attacking it. For instance, saying “could you explain why X implies Y, as I’m having trouble seeing the connection” rather than saying “X doesn’t imply Y, you idiot.” Generally, the more people feel attacked, the more they feel they need to defend themselves, and the more likely it is that discussions with them will devolve into fights. Another useful technique, if you’re having trouble making headway, is to rephrase your disagreement with them. If they claim “X or Y” must be true, rather than just saying “that isn’t right” for the third time, you can try “Can you think of any other possibilities besides X and Y? Because I’m not convinced that one of them has to be true.” Or if someone says “X implies Y” and you’re quite sure it doesn’t, you can ask “so if Y were not true, you think that would imply that X is not true?” This helps people look at their own claims from a different angle, which may make them realize problems in their arguments that they didn’t identify before.
When we encounter a complex and difficult to resolve disagreement, we can gain insight into why it is occurring and how it might be resolved by considering the roles of the seven (and particularly, the first five) fundamental reasons for disagreement: facts, definitions, values, signaling, failures of logic, information processing methods, and default beliefs.