Section 1: Introduction
To this point we have highlighted the importance of testimony, and discussed the conditions under which we should accept what other people tell us. However, we can look to other people for other kinds of information as well. In some cases, the mere fact that other people agree or disagree with us can tell us something important about the accuracy of our beliefs, and give us reason to update them accordingly. Moreover, we can look to groups of people for information. In these kinds of cases we take the fact that something is popular as evidence of its merit. In the right circumstances, and when the conclusion is drawn carefully, this can make perfect sense. In this chapter, we will look at these ways of learning from others, identify key questions to ask, and think about some of the factors that influence what we infer from other people’s beliefs.
Section 2: Agreement and Disagreement as Evidence
Recall our discussion of cooperative disagreement in Chapter 2. There we began by noting that disagreements naturally raise all kinds of good questions: Why don’t we agree? How could somebody see this differently? What are my reasons for thinking this in the first place? Am I missing anything? The goal of the discussion in Chapter 2 was to show the value of engaging cooperatively with others to understand disagreements, and to highlight some of the social and personal obstacles to doing so. As we saw, a disagreement can indicate that we’ve missed something or made a mistake, and can thereby present us with an opportunity to correct ourselves. Similarly, as we will see, the fact that somebody else agrees with us can bolster our confidence in our belief. In this section, we will set aside the issue of cooperative dialogue as a social practice to focus on agreement and disagreement as evidence. That is, we will focus on clarifying when, and under what conditions, agreement and disagreement tell us something about the accuracy of our own beliefs.
In order to illustrate how agreement and disagreement can serve as evidence, let us start with a couple of examples.
Imagine that you and a friend individually work through the same complex math problem. You get the answer ‘42’. Your friend has roughly the same math skills you do, but she got ’38.5’. Upon finding this out you are less confident your answer is correct, and go back to double-check it.
You are in the same situation, but when you check with your friend she says ‘42’ as well. Hearing that she agrees with you makes you even more confident your answer is correct, and you move on to the next one.
In these examples, your friend’s agreement/disagreement tells you something about your belief. Importantly, however, this is not always the case. Sometimes agreement and disagreement don’t suggest anything about the accuracy of your own beliefs. To illustrate, say that in Ex. 1 you methodically worked through the problem, but your friend was distracted and did it quickly and carelessly. Upon finding this out, your friend’s disagreement shouldn’t make you less confident, since you are more likely to have gotten it right than they are. Alternatively, imagine in Ex. 2 you find out that your friend did not actually work the problem, but instead looked at your answer, and reported it as her own. Upon finding this out, your friend’s agreement shouldn’t make you any more confident—after all, she is just repeating back your own answer. Consequently, the fact that somebody agrees or disagrees with us is informative in some cases but not others. What is the difference?
Let’s start with Ex. 1. In this case, it makes sense for you to be less confident about your answer, but why? On the one hand, you have diligently worked through the problem, and so you have good reason to believe the answer is ‘42’. On the other hand, in checking with your friend about the answer, you are treating her as a credible source for this problem. That is, you are taking for granted that she is being sincere and is in a good position to have figured out the problem correctly. Given that she is credible in this case, you have a reason to believe the answer is 38.5. But now you have conflicting pieces of evidence. Overall, then, the total evidence available to you now supports the answer ‘42’ less strongly than it did prior to hearing from your friend, and you should update your system of beliefs accordingly. In contrast, if your friend was distracted and careless in working the problem, then she may not be credible in this specific case (even if you think that under normal circumstances she would be). But if she doesn’t have credibility in this specific case, then the fact that she disagrees wouldn’t give you a reason to doubt your view.
Let’s say that your friend is credible in this particular case of disagreement. How should you update your system of beliefs? In general terms, this is going to depend on how you balance the conflicting pieces of evidence. First, how confident are you that you did the problem correctly in the first place? To put it differently, how confident are you that you didn’t make a mistake when you worked the problem? Second, how credible is your friend in this case? Thus, in updating your beliefs you need to compare the weight of these two pieces of evidence. For example, it might be that you were very confident in your belief that ‘42’ is the answer, and you persist in believing this is the correct answer even after you’ve taken into account your friend’s credibility (although with a lesser degree of confidence). That is, your reasons for thinking the answer is ‘42’ carry more weight than your reasons for thinking it is not. Alternatively, if your friend is very credible, and you weren’t very confident in your answer to begin with, then it can make sense to abandon your belief that it is ‘42’ for the moment, and say ‘I don’ know’. To push this example even further, imagine you find out that solving this kind of problem was part of your friend’s job last summer, and she has solved this kind of problem many times at work. In this case your friend is highly credible and her conclusion carries a lot of weight; if, in addition, you were not very confident in your answer in the first place, then your friend’s belief might fully outweigh your reasons for believing it is ‘42’, and tentatively accepting 38.5 would make sense. Importantly, regardless of how you ultimately update your belief, each of these cases is a good opportunity to pursue a cooperative dialogue and figure out what, exactly, accounts for this discrepancy so that you’ll both know better how to do the problem next time.
What about cases of agreement? As with disagreement, credibility is crucial. In Ex. 2, you have good reason to think the answer is ‘42’ since you’ve worked through the problem yourself. When your friend is credible, then the fact that she agrees with you gives you an independent reason to believe ‘42’ is correct. Consequently, the total evidence available to you at this point supports the answer ‘42’ more strongly than it did prior to hearing from your friend, and it makes sense to subsequently raise your confidence. If she is not credible, either because she doesn’t know how to do the problem or didn’t try very hard or was distracted, then there is no particular reason to think that she got it right. Consider a different case.
You are a big fan of NBA basketball. You watch a lot of games on TV, follow scores and statistics on a daily basis, collect basketball cards, and so on, and you’ve done so for years. One day at lunch, you say, “I think that San Antonio will be one of the best teams in the league this year”. One of your friends agrees, saying “yeah, I think they are going to be really good.” However, you know that your friend grew up in San Antonio. He doesn’t really follow basketball, but is always optimistic about his home team—he thinks they are going to be good every year.
In this case, the fact that your friend agrees with you doesn’t give you any additional reason to think San Antonio will be very good this year, since his belief is not based on a consideration of any of the relevant evidence for that claim (the relative strengths and weakness of the team, for example). It is it important to make two points about this. First, notice that had your friend disagreed with you, it shouldn’t lead you to change your mind either. Your friend’s lack of credibility on this issue means that their view shouldn’t lead you to update your beliefs in any way, since you know a lot more about the issue than they do. Second, unlike disagreement, in cases of credible agreement you do not need to balance competing weights, you need only add to your existing confidence.
Overall, that fact that somebody else agrees or disagrees with us can tell us something noteworthy about our own beliefs. The key issue to determine in these cases is whether the agreement/disagreement is coming from somebody who is credible on the claim at issue.
A Question to ask about Cases of Agreement or Disagreement:
- Is the other person credible about this specific claim?
When others agree or disagree with us, that agreement or disagreement matters only when it is credible. That is, we should update our beliefs in light of their agreement or disagreement only when the person is in a good position to have an accurate belief in this case. One important question remains: what if you don’t know whether the other person is credible or not? In this case, you can’t say whether they are likely to be correct or not, and so there is no reason to update your beliefs. Nevertheless, like the other cases, this can be an opportunity for open-minded evidence gathering and cooperative dialogue.
Section 3: Appeals to Popularity
A different way that we use other people to guide our thinking is called an Appeal to Popularity An Appeal to Popularity is an argument that draws a conclusion about the truth or merit of some claim, behavior, or product on the basis of its popularity. Consider the following examples:
It’s okay to photoshop pictures of myself and others and post them online, everybody is doing it.
I think the change would be great for the sport. After all, over 80% of players favor the change.
Appeals to popularity can be subtle; sometimes the conclusion is not explicitly stated but is merely suggested. Consider the following case:
Ford Trucks: the best-selling trucks in America.
On its face this is just a claim about Ford’s sales figures, but Ford is not just telling you about their sales because they think you might be interested. Rather, they are hoping that you will take this fact as indicative of something; they hope you will reason as follows:
- Ford has the best-selling trucks in America
- So, Ford trucks are good in some important way (reliable, high quality, etc).
Indeed, Ford would probably like you to go even further and conclude that Ford has the best trucks in America.
In calling this kind of argument an Appeal to Popularity, it is important to note that ‘popularity’ is being used broadly to capture a variety of different notions. In Ex. 4, the author says that “everybody is doing it”, but they probably don’t mean that literally. What they probably mean is something like this: “many people I know of are doing it” or “it is popular among many of my friends”. In contrast, Ex. 5 is more specific. In this case, the proposal is popular among players of the sport. Thus, the first point is that saying a belief, proposal, or product is popular is really saying that it is popular among a particular group of people. A second point is that a belief, proposal, or product may be popular to different degrees, and to say that something is popular is not necessarily to say that it is popular among all or most people in the group. Take Ex. 6: in pointing out that Ford has the best-selling truck, they are not thereby saying that all or most Americans have bought a Ford Truck. Rather they are making a relative claim—Ford sells more trucks than other companies. Overall, then, the term ‘popularity’ is not very precise in this contexts.
Keeping this in mind, let us lay out the basic structure of an Appeal to Popularity as follows:
- x is popular among the members of some group.
- So, x is probably good in some respect.
Put in these terms, we can see that there is a gap in this argument. What does the fact that something is popular have to do with its truth or merit? It can be easy to miss, but arguments like this are actually instances of Inference to the Best Explanation! An Appeal to Popularity begins with an observation that some subject is popular, and then jumps to a specific explanation for its popularity, namely that it is good. As a result, we can use what we know about Inferences to the Best Explanation as a guide for thinking about Appeals to Popularity. Thus, we can characterize the missing premises in Appeals to Popularity accordingly:
- x is popular among the members of some group.
- The most likely explanation for why x is popular among members of this group is that x is good in some respect. (MP)
- So, x is good in some respect.
Setting out the argument in this way is especially helpful for understanding the difference between logically strong and logically weak Appeals to Popularity. However, before we turn to this issue we will briefly consider what motivates us to use this form of argument more generally.
Section 4: The Appeal of Appeals to Popularity
As we will see, Appeals to Popularity are often logically weak. Nevertheless, they are common in everyday thinking, and so it is worth taking a moment to think about why we use Appeals to Popularity. Perhaps the most straightforward answer is that they are easy short-cuts. Go back to Ex. 6 above. Suppose that you draw the inference that Ford hopes, and come to the conclusion that Ford Trucks are probably pretty good trucks, since they are the best-selling ones. Think about your evidence here. In coming to this conclusion, you didn’t look at any of the factors that might actually make a truck good in some way. You didn’t look at its reliability, its towing capacity, its gas mileage, its acceleration, the size of the truck bed, and so forth. Instead, you’ve looked at what people believe about the truck (namely that it is worth buying). This is a lot easier—especially when Ford tells you about it—than doing the work for yourself. Many Appeals to Popularity are like this: we use what is popular as a shortcut for identifying what is probably true, or good, or meritorious in some way. In general, when popularity is genuinely an indicator of truth or merit, there is no problem with taking this shortcut, though again, this is not common. Moreover, even when Appeals to Popularity are sound, they do not help us understand why something has merit or is true, only that it does. Consequently, when it is important to understand something for ourselves, an Appeals to Popularity won’t suffice.
Appeals to Popularity are easy, but there are other factors pushing us to look to what is popular as a guide for our belief and action, namely an interest in conformity. In a sense, this is probably not too surprising; after all, everybody has experienced peer-pressure in one way or another. What is, perhaps, more surprising is the power these forces can have over us. In a well-known series of experiments psychologist Solomon Asch sequentially presented a group of participants with two cards. On the first card was a single line. The second card had three lines on it.
The researchers asked each individual in the group to publicly judge which of the three lines was of the same length as the line of the first card. This is an easy task, the answer is obvious, and the first few times this task was completed everyone agreed about the obvious correct answer. What makes this experiment interesting is that all the participants–but one–were part of the experiment and were acting off of a script. They put this to work in the next trial: the first actor gave an obviously wrong answer, and all the other actors agreed. Researchers found that in about a third of the trials the last participant would conform to the rest of the group and give the obviously wrong answer. These kinds of experiments have been conducted in a variety of ways, but the result remains the same. In the face of a unanimous—but obviously false opinion—people will surprisingly often act in conformity with the group, as opposed to in conformity with the truth. Given the motive to conform, it is not surprising that we often use the evidence of other’s beliefs and choices as a guide for our own.
Section 5: Evaluating Appeals to Popularity
When is an Appeal to Popularity logically strong? Given that Appeals to Popularity are instances of Inference to the Best Explanation, we can follow the same guidelines. Thus, as we saw in Chapter 13 step one is to ask whether the proposal that the subject is good in some respect would plausibly explain its popularity. In general terms, yes. We prefer products and proposals that are good in some way, and we prefer beliefs that are true. Let’s turn to step two. Step two will be to look for plausible alternative explanations for the popularity of the subject. As we know, there are many possible explanations for any state of affairs. Popularity is no different. Take a product for example. It might be the most popular because it is the best advertised, because it is the cheapest, or because it is carried by every store, and so on. As an illustration consider this case:
McDougal’s sells more cheeseburgers than any other company in the U.S. So, McDougal’s cheeseburgers must be the best.
The fact that McDougal’s cheeseburgers are the most popular cheeseburger in the U.S is a consequence of a number of factors which surely include price and easy access to McDougal’s stores. This brings us to step three. Now that we have some competing explanations, we have to compare them and see if the proposed explanation comes out on top. In the McDougal’s example, it does not. That is, the facts that McDougal’s cheeseburgers are inexpensive and are easily available to millions of people seem like a better explanation for the popularity of McDougal’s cheeseburgers than the proposal that they are the best.
The argument in Ex. 7 is a logically weak Appeal to Popularity, but what would a logically strong one look like?
A logically strong Appeal to Popularity would be a case in which a belief, proposal, or product’s popularity really is best explained by its merit. So, when it this the case? For sake of ease, let us talk simply in terms of a product. A product’s popularity would be best explained by its merit if it is popular precisely because many people investigated the relative merits of the product, and on that basis, decided to purchase it. Thus, when we are looking for logically strong Appeals to Popularity, we need to look for cases where the best explanation for a subject’s popularity lies in many people’s thoughtful determination of the subject’s merit. As you can probably see, this is not common. Nevertheless, there are some logically strong Appeals to Popularity in our everyday experience. Let’s take a look at one.
Section 6: Popularity Among Experts
There are some cases in which something is popular as a result of individual investigations of merit. Perhaps the most common case has to do with populations of experts. Consider the following example.
Delia: I don’t get why people are so concerned if there are fewer bees. I mean, they sting you. The less of them the better.
Sam: Biologists agree that bees are a crucial part of the ecosystem because of their role in pollination.
Sam is not just reporting what one Biologist says, but is using the fact that there is consensus among biologists on this issue. This is a logically strong Appeal to Popularity because biologists are experts who in their professional work seek to make judgments about what is true on the basis of the available evidence. More broadly, when there is consensus about some thesis within the community or population of experts, and that thesis is within the community of experts’ field of expertise, then a logically strong Appeal to Popularity can be drawn.
As another example, consider the fact that some of the current debate about global warming is actually a debate about whether a logically strong Appeal to Popularity can be drawn by the public. Most people are not in a position to personally evaluate the evidence for anthropogenic climate change (that is, the thesis that climate change is occurring and that it is caused by humans). After all, the vast majority of people do not have access to relevant data sets, they do not have the requisite statistical/mathematical expertise, they do not have the background knowledge required to read and assess the literature, and last, they do not have the time. For these reasons, most people’s beliefs about anthropogenic climate change are based on the limited evidence they can personally evaluate and on their knowledge of what the population of experts believe. In recent years a number of studies have found that well over 90% of climate scientists believe in anthropogenic climate change. Given the criteria we have discussed, it would seem that this fact offers good evidence for thinking that anthropogenic climate change is really occurring—after all, there seems to be consensus on the issue from experts in the field.
The idea of appealing to a population of experts is not limited to reasoning about science. Consider the well-known website rottentomatoes.com. The chief function of this site is to aggregate movie reviews. For each movie that is currently in theaters (and for some that are coming soon) the site offers a measure on the “tomatometer” scale. This is just a measure of how many critics have positively reviewed a movie. Thus, if 75% of the available reviews positively evaluate a movie, it has a rating of 75. This information is intended to help the movie-going public decide what movies to see. The underlying idea is that people who write movie reviews are (or tend to be) experts, and since popularity among experts is indicative in some sense of quality, the tomatometer can be a useful guide for making movie-going decisions.
Section 7: What Does Popularity Tell Us?
Appeals to popularity are not often logically strong. There are many reasons a belief, behavior, or product might be popular other than that each member of the population investigated the belief, behavior, or product in question and determined it to be meritorious, and when compared to these alternatives merit is rarely the best explanation. Nevertheless, it seems like popularity often tells us something—especially when it comes to products. Consider the following situation: you are visiting a new city and have been staying in a hotel for a few days. Outside your hotel you have noticed that there are two similar restaurants. Both restaurants offer pub-style food and are located right across the street from one another. You have noticed that one of these restaurants is consistently busy, while the other is consistently dead. One evening you decide to eat near your hotel and these two restaurants are your only options. Which restaurant do you choose?
Most people choose the busy restaurant. But does this choice make sense? After all, this choice relies on an Appeal to Popularity and we have learned that there are many reasons a restaurant might be popular other than that the food is good—it might have been doing a lot of advertising, it might have really good drink specials, it might have attractive waiters/waitresses, it might be exceptionally cheap, etc. This choice does make sense, but not because it will likely have good food. The choice makes sense, rather, because it is likely that the food is not immediately and obviously poor (whereas you have no such evidence for the dead restaurant). At least when it comes to products, the fact that it is popular over time typically offers evidence that it isn’t obviously awful or a clear failure, since products that are obviously awful or clearly fail to do what they claim are not popular over time. This is defeasible evidence, of course, and this is no guarantee, but it is evidence nonetheless. Consider Ex. 6 again. Although we cannot infer from the fact that Ford Trucks are the bestselling trucks that Ford’s trucks are the best, we can reasonably infer that they are not immediately and obviously bad (of course that is not exactly the inference that Ford wants you to draw—Ford Trucks: Not Awful).
Section 8: The Psychology of Popularity Judgments
Using an appeal to popularity presupposes a belief that some idea, behavior, or product is popular; but how do we know whether something is popular? In some cases our belief about something’s popularity among a specific group of people is based on solid statistical data, but most of the time it isn’t. In fact, most of the time we simply rely on our intuitive sense that something is common or uncommon, frequent or infrequent. When making these kinds of judgments we need to be careful since psychologists have shown that we tend to overestimate the popularity of beliefs we share in, and underestimate when we do not. This phenomenon is called false consensus. As the social psychologist Ziva Kunda writes:
False consensus has been shown to color people’s estimates of the prevalence of just about any choice, attitude, or behavior that has been examined, in domains as diverse as everyday habits and preferences, personality traits, and political opinions. For example, fans of white bread think more people would choose white bread over brown bread than do fans of brown bread, optimists believe optimism is more common than do pessimists, Americans who prefer the Republican presidential candidate believe that support for the Republican is more widespread than do those who prefer the Democratic candidate, and those who support cuts in public spending believe that more people support such cuts than do those who oppose them.
To be clear, to say that we have this tendency is not to say that we always overestimate the relative frequency that others share our beliefs and attitudes. Nonetheless, this is a common phenomenon, and since appeals to popularity rely on a judgment that something is frequent or common, it is something we should be aware of in thinking about appeals to popularity.
Overall, Appeals to Popularity are common, and it is not hard to see why. They are an easy shortcut for determining subject’s merit. Nevertheless, they are logically strong only in cases where merit is the best explanation of a subject’s popularity. We find this in cases where experts agree, but in general, logically strong Appeal to Popularity are few and far between. Moreover, the phenomenon of false consensus shows us that we are subject to mistakes in judging popularity in the first place. The upshot here is that we should be extremely wary of Appeals to Popularity in our everyday thinking. In particular, we need to remember the following:
Two Questions to Ask of Appeals to Popularity
- Are there other plausible explanations for the subject’s popularity?
- Would the truth of the proposed explanation be less surprising than the truth of any competitor?
Exercise Set 18A:
Directions: Consider the following cases of agreement/disagreement. In each case comment on whether the agreement/disagreement in question implies that you should update your confidence or even change your belief. Explain your answers.
You think the midterm is in two weeks, since that is what it says on the syllabus. You overhear two people from class saying that the midterm is next week.
You are sitting in class, and it sounds like the teacher has just said the midterm was moved up. But somebody coughed at the same time the teacher said this, so you wonder whether perhaps you misheard. You turn to the student next to you and ask, “did she just say moved up?” The student says ‘yes’.
At lunch one day, the topic of CPR comes up and somebody asks which you should do first when you give CPR, rescue breaths or chest compressions. Because you had to complete a course in CPR and be proficient for your summer job as a lifeguard you know the answer, and say ‘chest compressions’. The person sitting next to you disagrees, saying, “no, in our school’s production of Wit we did rescue breaths first.”
You are reading a magazine and there is an interview with a chef from Santa Fe, New Mexico. In the article the chef mentions the challenge of cooking in Santa Fe where water boils at less than 200 F. This strikes you as wrong. You learned in school that water boils at 212 F and water is pretty much the same anywhere you go.
You are catching a ride with a friend to a city across the state for a wedding. You think route A is the fastest route because that’s what your phone’s mapping app says, but she thinks Route B would be quicker because that’s what her phone’s mapping app says.
Exercise Set 18B:
Directions: Determine whether each of the following is an Appeal to Popularity or not. If not, try to identify the argument’s form. Then comment on the argument’s logical strength.
Most small business owners agree that they pay too many taxes, so the legislature should take tax increases on small businesses off the table.
Evaluate B’s argument
A: I guess I just don’t think hunting is ok—we don’t need to do it, and I don’t think killing something for sport is acceptable
B: But c’mon. People all over the world enjoy hunting, and have been doing so for thousands of years.
You have recently moved to a new neighborhood and don’t know what day you should put out your garbage. You decide to wait and see what everybody else does. You see that on Wednesday night all your neighbors put out their garbage, so you decide to do so too.
A: I think we should just buy the new truck and call it a business expense so we can write it off on our taxes.
B: I don’t know. That sounds like cheating to me. We wouldn’t really use the truck very much in the business, you know.
A: Oh, don’t worry about it. This kind of thing is done all the time.
9/10 podiatrists recommend against wearing the same pair of shoes every day, so I guess I should add some variety to my shoe wearing.
There are very good reasons for the death penalty. First it serves as a deterrent to those who would commit capital offenses. Second, it is just and fair punishment for the crime committed. Third, reliable opinion polls show that over 70 percent of all Americans favor it. If so many people favor it, it has to be right.
Exercise Set 18C:
In the Asch conformity experiment discussed in Section 4 above, one of the reasons that people gave for conforming in spite of the obvious evidence is that they started to doubt their own belief. They reasoned that everybody else must have known something they didn’t. How does this connect to our discussion of disagreement? What question should people in this experiment have asked themselves, and how should they have subsequently evaluated the situation?
What about unpopularity? Is unpopularity ever a good reason to think that something is false or poor? When?
Suppose that you are an unscrupulous CEO of a corporation that makes product x, and there is consensus among scientists that your chief product is toxic. It seems like it is only a matter of time before your product is banned by the government. How might you try to manipulate public opinion about your product to prevent its being banned?