Chapter 20 Putting Critical Thinking into Practice

20.1 Returning to Inductive Strength

The nature of a critical thinking text such as this one is to show you many ways in which arguments go wrong—we show far more ways to go wrong than to go right. This might lead to a feeling that there are logical booby traps everywhere and making good arguments is impossible. This is not what we would like you to take from the text. The chapters on deductive reasoning showed that necessarily true conclusions are possible with good form and true premises. In the real world, we are often making arguments that do not fit neatly into this form. However, there is still a lot of reasoning we can do that gets us toward the truth.

The fact that most knowledge is merely probable rather than certain is obvious; most of our predictions are based on estimates or fallible signs of things to come; for example, we rely on the weather report even though we know well that it may be wrong. These days, it isn’t uncommon to open the weather app and read that “it is snowing” in your location when it is indeed not snowing in your location. This is just a baked-in feature of inductive reasoning: it’s probablistic and vulnerable to new information. More information will not dissolve this vulnerability. Inductive reasoning is limited because human beings have intellectual limits (we are not omniscient) and we can only live in the present (we cannot know the future). The human condition is one of incomplete knowledge in a world where we have risks and opportunities.

So, we are limited but must do our best with what we have. Our intellect is composed of many fast processes that run as automatic problem-solving modules that work reasonably well in certain situations. Like other animals, we evolved inferential powers to help us navigate our world. If your cat comes running whenever you open a can, we can say she has learned to recognize the sound of the can opener as a sign of probable food. Even if it is cat food only a quarter of the time you open a can, your cat thinks it is worthwhile to check and see. She is able to adjust her behaviour based on a recurring but low probability.

Many of the inferences we make are of just this sort: they are fast, very fallible intuitive judgements that rely on typical features, probable signs, and reasonable assumptions and that allow us to avoid risk and pursue goods. While they are individually more like guesses than knowledge, they cohere with one another—some being ruled out by the joint testimony of others—to give us an overall reasonable and reliable sense of what is going on around us. These assumptions we make about the world taken together are very powerful. A critical thinking course gives us a chance to reflect on these assumptions and pictures that we build of how the world works.

From small everyday assumptions to metaphysical claims, the tools of critical thinking remain the same. What is the justification for a claim? Is it necessary? What is the evidence? What supports the evidence? Even going to the store to buy milk contains a large number of assumptions, reinforced by other assumptions; the store will have milk, we can afford it, etc. More generally, we operate on assumptions that can also be scrutinized, such as the existence of money, what it does, what a store is, the existence of cows, that milk is a food, that what is in the carton is milk, and so forth. Our “merely probable” beliefs are enmeshed in a network of support relations with other beliefs. Again, this might start to feel like everything is questionable and you can’t know anything. But let’s follow that line of thinking. The fact that your beliefs are merely probable doesn’t mean that you can just believe whatever you like.

This exercise only works if you don’t just imagine it to be true but really try to believe it (that you have a Ferrari or that your parents live in another country than they do or that you are eight feet tall). Think of all the consequences the truth of such a statement would have for your actual day-to-day life, in which the statement is shown to be false by all the other things you know. Of course, you may be able to imagine that the statement is true, but you cannot actually believe it at will. Of this, you are certain. Consider, then, how our “merely probable” beliefs are not only enmeshed in a network of support relations with other beliefs; they fit together into patterns of meaningfulness that make them accessible to our understanding. Our general knowledge base is not simply a large list of beliefs to which we give assent. Instead, our general knowledge is organized into various kinds of patterns that make our beliefs relevant and accessible.

Here are some examples of how we organize beliefs: in stories, plots, goal-oriented plans, schemas and scripts, or stereotypical and casually regular situations. These function as organizing structures that package our judgements into useful patterns for our lives—plans and activities that we pursue. Think about these patterns: going to a restaurant, playing a game, going on a holiday, getting married, solving a problem. Imagine that you are playing a game of improvisational theatre at a party, and you are told to act out “going to a restaurant.” Whether you would be good at acting or not, you know what to do—after all, you know what to do when you actually do go to a restaurant. It is also useful to reflect on how we understand the behaviours of pets or of small children and how their knowledge is organized, since they show simpler forms of our own patterns.

The point is that we make many kinds of implicit inferences that do not involve definitional truth but rather typical features that “everyone knows” or probable consequences based on function or a likely purpose or goal. Most human stories are plan-based stories; information is organized around people in particular situations who perform understandable actions for reasons that are available to anybody. The inferences that depend on these structures may be called material inferences, since they do not depend upon a formal pattern, but instead on an informal pattern that is neither universal nor necessary but is useful and productive. It is a general characteristic of material inferences that they can be defeated by additional information; they are “defeasible” or fallible.

Material inferences are always potentially vulnerable to more information: it rains, and so plans for a picnic get changed; you get offered a job, and your summer plans shift to accommodate it. Material inferences are both strong and weak. Their strength is that they permit you to form a conclusion that you can act on with the information you have at hand; it allows you to assume that things are normal and will go as you expect. Here is an example: If we tell you that we are going to the store, you may conclude (in the absence of further information) that we want to buy something. But if we now tell you that we have promised Fred a ride home and that Fred is at the store, then you would probably no longer conclude that we want to buy something. But if you heard us tell Fred that we would give him a ride because we need to pick up some things at the store anyway, then you could after all conclude that we want to buy something. Of course, if you know that we acquire goods by shoplifting, you will not be able to conclude that we want to buy something . . . and so on. Each additional bit of information has the potential to change what you will infer. Material inferences are a kind of enthymeme, and it is their pervasiveness that leads some logicians to argue that material inferences are ineliminable in human reasoning (we can’t not use them all of the time). Here are some examples of different types of ordinary material inferences.

People have motives for their actions, and their actions are organized into plans that are guided by purposes. We understand a person’s behaviour by recognizing both what kind of action it is and what kind of motive would explain why the person did it. Since people can have many motives, our inferences are easy to overturn. However, when we speak to one another, we tend to give people salient information—that is, relevant information that will make it easy for them to infer what we wish to convey. So if all we are told is that someone did a stereotypical action (Bill went to the store), we will infer that they were moved by the likely motive (he wanted to buy something). Motivational assumptions are potentially risky; people may have unusual motivations that you don’t know about, they may lie to you or attempt to swindle you and so mislead you about their motives, but these people are truly rare. Motivational assumptions are unavoidable, and we make them constantly because they make other people and their actions intelligible to us. We also make another kind of material inference: a feature inference.

So it is an inference from a stereotypical property to the bearer of that property. Babies typically wear diapers (but of course, so do incontinent adults and others). If we tell you that we need diapers for Andy, and we don’t tell you anything else, you assume reasonably that Andy is a baby because diapers are stereotypically used by babies, and most diapers that are used are used by babies. Additional information can block the inference; if you know that we have a sick and aged poodle named Andy, then you will not infer that Andy is a baby. Another material inference is a resultative inference.

If you ask why Fred didn’t come to the movie, and we tell you Fred hit his head, you will infer that Fred was injured and that his injury explains his absence. Obviously, such an inference can be defeated by additional information. Resultative inferences are required in ordinary prediction; you step on the gas because you believe that this will make your car move, for example. We also make inferences about the function of something.

Hammers are for hammering, chairs are for sitting on, food is for eating, and so on. If we ask for a hammer, you rightly infer we wish to hammer something, or to obtain a hammer for someone who wishes to hammer something.

Material inferences of this sort are central to language use and to successful communication; after all, they are based on patterns that we all use and take for granted. We saw in our discussion of ambiguity (Chapter 14) how understanding ordinary sentences is a function of both weak grammatical rules and background information that can overrule one interpretation in light of something one knows. Because information is always limited, we normally expect you to organize the information that you communicate to us so that it is easy for us to understand you. Furthermore, you will normally give us the information we need to understand what you are saying, and because you do this, you have a right to expect us to understand your point. The misuse of material inferences, whether deliberately or by accident, is a common cause of fallacious informal reasoning (as we saw especially with analogies, fallacies of false cause, and our discussion of generalizations). However, as long as we treat them as provisional and stay open to dialogue (the giving and receiving of reasons), they are very useful in our day-to-day lives.

20.2 Making Better Arguments

By now you should be able to identify many of the ways that ordinary reasoning can go wrong. We have seen that these really represent diagnostic categories and that there is sometimes more than one way that a piece of reasoning can go wrong at once. This is especially true in more extended arguments where there are many opportunities for one kind of error to produce another. Just as a doctor will see certain bodily signs as symptoms of a disease rather than the disease itself, the careful reasoner will look at fallacies as symptoms of bad reasoning that will guide in both analysis and cure.

In part 3, we defined a fallacious argument partially as an argument that is not cogent and in part 1 we defined a cogent argument as one that meets three conditions:

These three conditions offer us the beginnings of a diagnostic procedure for evaluating extended arguments. To apply the first condition, we must first of all identify all the claims being put forward and distinguish the conclusion from the premises so as to highlight the reasoning between them. As we have seen in the discussion of the fallacies, what is presented as the conclusion and what is actually being argued for are not always the same. There may be lexical ambiguity or irrelevant thesis, or the premises may be designed to move our emotions rather than present reasons. Once we have charitably identified all the claims put forward and identified the conclusion from the premises, we are in a position to determine whether the premises are dialectically acceptable to a reasonable audience. Recall that to be dialectically acceptable, the premises must not only be true or likely but also appropriate to support the conclusion. If the argument is circular, then even if the premises are true, they cannot offer rational support to the conclusion; if there is ambiguity between premise and conclusion, then the appearance of a support relation will be illusory as well. If the argument is neither circular nor ambiguous and the premises are otherwise dialectically acceptable, we can next ask whether the premises make a rationally grounded connection to the conclusion. The fallacies of emotional bias in particular fail this test. However, to say that there is a rational, grounded connection doesn’t yet show that the connection is strong enough to allow the mind to move from the acceptability of the premises to endorsing the conclusion. The fallacies of expertise and the inductive fallacies can be used as a kind of checklist for determining the strength of the connection.

The fallacy method of critical thinking involves identifying claims, distinguishing premises from conclusions, and asking whether premises are dialectically acceptable. If they are, we then investigate the kind of connection the premises make to the conclusion: If the connection is not genuinely based on reason, we reject the argument; if it is reason based, we ask whether the ground of support is rationally sufficient. At each step, we can use our fallacy list of “bad argument patterns” as diagnostic tools. Along the way, we also bring to bear considerations of good practice that we have identified in the text. We look at the questions carefully and methodically. If there are implicit premises about what words mean or about what everyone knows, we try to make those assumptions explicit. If appeals are being made to our interests or desires, we step back from them to see whether the appeals are legitimate or whether they simply attempt to influence our judgment.

Recall that in addition to a fallacy lacking cogency, we also defined fallacies in terms of Douglas N. Walton’s five conditions for defining fallacies (1995, 255):

Fallacies have many features that make belief tempting—all the more reason to learn to identify them in our everyday reasoning.

20.3 Evaluating Arguments in Longer Text

Consider the following excerpt from a letter to the editor in the St. Albert Gazette (Letter to the Editor, “Don’t Cut Money for Seniors,”1 February 13, 2023):

In a recent letter to the editor (Feb 2023) a resident wrote in asking that council change its plan to cut funding for services for seniors. Here’s an excerpt from the piece:

“Discussions with other seniors has indicated to me the lack of common sense council used to spend over a million dollars on a traffic circle for marginal benefits while ignoring the needs of seniors who have contributed a lifetime of revenues to the city. My father was a federal, provincial and municipal politician who was a president of the Alberta Urban Municipalities Association (AUMA), was president of the Alberta Municipal Service Corporation and was named one of Alberta’s ‘50 Most Influential Citizens’ by Alberta Venture magazine and specifically identified as one of seven ‘political power houses’ in the province. His lifetime as a politician over 50 years; he profoundly indicated the need to be aware of and communicate with seniors as to what their needs were. Seniors have always been considered to be the electoral base. When is St. Albert council and the UCP going to communicate with the seniors and better understand our needs?”

Looking at these words, you see a passion for seniors and a strong desire for money to be spent in a responsible way for seniors. We might feel this way too, but do the stated premises support this? We can extract a number of sub-arguments that contain fallacies. Let’s first deal with this part of the passage:

Discussions with other seniors has indicated to me the lack of common sense council used to spend over a million dollars on a traffic circle for marginal benefits while ignoring the needs of seniors who have contributed a lifetime of revenues to the city.

This one sentence contains a number of issues that we can evaluate. First, when the author says “discussions with seniors,” is this an appeal to an anonymous authority? It seems to be, since they have not cited who they talked to, how many people they talked to, or whether they were experts on what they are being cited about, which is that city council lacks common sense.

Next, saying that the council lacked common sense is abuse. It is a way of glossing over calling them a worse name; however it is not nice to say someone lacks common sense. Essentially, “common sense” is used to undermine the argument by drawing attention to the person, not the claim—in this case, the claim is that a traffic circle needed to be built (at the time).

So what about that traffic circle? Even a small town like St. Albert (about sixty-six thousand residents) has a multi-million-dollar budget. Looking at the financials for 2023, it is hard to ascertain what the total spending amount is available for a city, given that a city would have investments, revenue-generating projects, and, notably, tax collection. Let’s just say, though, that a million dollars on a traffic circle (if that’s true, and it seems to be) is a lot of money, but it is a small proportion of the overall budget. And notably, the arguer is bringing this up in the context of proposed funding cuts for seniors in 2023 when the traffic circle was completed in 2015. This means that money is long spent, not to mention members of council have changed, and the cuts to senior funding are proposed in 2023—a different pool of money. So is this a fallacy?

We hope by now you can see that this is an irrelevant thesis. The fact of purchasing a traffic circle (even if it was not a good idea and even if it was more expensive than it should have been) is not relevant to the amount of the budget spent on senior services, especially eight or more years later. Consider the following possible reconstruction:

• Premise 1: City council lacks common sense.
• Premise 2: City council spent one million dollars on a traffic circle.
• Conclusion: Therefore, do not cut money for senior services.

Spending one million dollars on a traffic circle is a fact, and it is a fact worth discussing, but it is not relevant to the issue at hand. Formulated in another way, you can see this is also a version of a straw person, since it characterizes council as a bunch of senseless nitwits who throw money at silly infrastructure while ignoring what is important. The idea here is that in reconstructing this argument in standard form, you can see how the premises provide no ground for the conclusion since they are irrelevant, abusive, and they don’t tell us anything about services for seniors.

Let’s consider the next section of the argument:

My father was a federal, provincial and municipal politician who was a president of the Alberta Urban Municipalities Association (AUMA), was president of the Alberta Municipal Service Corporation and was named one of Alberta’s “50 Most Influential Citizens” by Alberta Venture magazine and specifically identified as one of seven “political power houses” in the province. His lifetime as a politician over fifty years; he profoundly indicated the need to be aware of and communicate with seniors as to what their needs were.

To summarise this and put it in standard form might already identify the fallacy.

• Premise 1: My father was a politician with extensive service and accolades.
• Premise 2: He said we should listen to seniors.
• Conclusion: Therefore, city council should listen to seniors.

First of all, it is already an issue that one’s influence (political powerhouse, influential citizen, etc.) is being conflated with expertise. This smacks of equivocation, since the two are being used as if they mean the same thing. It is possible that none of this person’s political service could have anything to do with being an expert on seniors’ issues. Based on this specific appeal to the politician’s experience, this is an argument that city council must listen to seniors (and presumably not cut their funding because that is what seniors would say). And since this is an argument that claims the conclusion is forced by virtue of what one person said, it is an appeal to authority. Is the authority genuine?

We can’t really evaluate the authority being appealed to, since they are not named. Maybe they have expertise and credentials in the area, maybe they don’t. Since we don’t know, we don’t have reason to accept the conclusion. In cases like this, it helps to imagine what genuine authority here would look like. It would have to be a collaborative endeavour of experts who have appropriately obtained representative information on what seniors in St. Albert want, how money can be spent in the most effective way, and a tremendous amount of information about the budget. We can get a little more information by reading on:

His lifetime as a politician over 50 years; he profoundly indicated the need to be aware of and communicate with seniors as to what their needs were. Seniors have always been considered to be the electoral base. When is St. Albert council and the UCP going to communicate with the seniors and better understand our needs?

It seems like he is threatening that if you don’t listen to seniors, you will be voted out. He does this by claiming seniors are the voting base, so it would be helpful to see if he can make good on this threat. We could start by asking whether being a “base” is just about numbers. About 75 percent of the population can vote, and 20 percent of the population is over sixty-five years of age. This means that 26 percent of the voting population is over sixty-five. And even if you correct for voter turnout among age groups, everyone eighteen to sixty-five represents considerably more votes than those over sixty-five (approximately 82 percent voting rates over sixty-five represents at most 21 percent of votes). This is not to say they aren’t an important voting bloc, but it grossly exaggerates the electoral importance of seniors’ votes. This is not to say that if you are a smaller voting group the majority ought to steamroll the minority. But this argument distorts the facts of how much influence seniors have in elections. Also, we’re guessing seniors have differences of opinion about who to vote for!

Not only is this an appeal to force or fear, but it is implicitly a bifurcation: “Do what we want or we will vote you out!” Politically, this might make sense—we do have a right to vote out those who are not representing us properly. However, we are in the business of evaluating arguments, not politics. Here, we need an argument about listening to seniors and not cutting their funding that is rationally cogent.

One area that we didn’t cover was the argument that seniors have contributed a “lifetime of revenues” to the city. This cuts to the heart of the purpose of city council and how it relates to seniors. Consider the following argument that we have constructed that might support this conclusion:

• Premise 1: Municipal governments collect taxes to provide support services to residents.
• Premise 2: Fair taxation involves redistribution of monies according to human needs, vulnerabilities, future planning, education, public good, and so on.
• Premise 3: Seniors require services that other citizens do not.
• Conclusion: It would be wrong to cut funding to seniors below the amount required to meet their needs.

We’re keeping the conclusion that it is wrong to cut funding rather than that council must listen to seniors because these are different issues. Here we appeal to the purpose of taxes (functional inference), then we appeal to a feature of taxation—that it is used to provide services (feature inference)—and then a fact about seniors’ distinct needs. Putting these claims together, we have a stronger argument that seniors’ funding should not go below what it takes to meet their distinct needs, within reason. We’re hoping this demonstrates that while the author has a good idea (it is a good idea to support seniors), they have not provided a cogent argument to support their conclusion. It is possible to provide better support than fallacious reasoning.

Extended pieces of reasoning may be too complex to consider all at once, but they will contain parts that can be isolated and evaluated independently. Throughout the evaluation of an argument, we need to recognize that we are not simply following some rules but that we are actively exercising our judgment and taking responsibility for the claims and connections being made. We need to be charitable, identify claims correctly, and rank the importance of statements in the overall pursuit of an argument. We hope to have demonstrated that you have to discriminate the pattern of argument and analyze the claims and their relationships in order to discern the fallacies they might contain. And in so doing, you might find more than one fallacy, but you have to decide which is most crucial to identify based on the features of the argument. There might be stronger and weaker versions of a fallacy, and you want to identify the strongest first.

Another way to say this is when we are identifying fallacies and evaluating reasoning in general, we are taking cognitive ownership of the argument as our own and thus as anyone’s. We take it on and pay very close attention to it, and we have to be charitable. The letter writer about seniors’ cuts is probably very irritated by the actions of city council. We should not just dismiss them as cranky jerks—we understand that they are likely exaggerating, but we still take them to be people making claims that can be evaluated for reasonableness. Also, we should be asking ourselves if there is anything good in the passage. Even if there are several fallacies, are there points worth considering? If we isolate the fallacies, can other parts of the argument survive intact?