1705136447
17.2 Argument by Generalization
1705136448
1705136449
An argument by generalization can be used to describe an entire group by presenting evidence from specific cases selected from within that group. Such an argument almost always relies on examples as evidence (discussed in the previous chapter). Generalization is based on the probability that examples selected from a group are likely to exhibit many of the same features of the group as a whole. In other words, by examining a representative sample of a group, one is able to make a statement about the group as a whole. The assumption is that characteristics observed in the group probably belong to not only the sample, but also to the group as a whole.
1705136450
1705136451
Thus, an argument by generalization moves from evidence about specific examples to a claim regarding the group as a whole. For instance, a debater wanting to argue that student athletes will become successful business people should begin by examining examples of students who were athletes in school and who later became successful in business. To construct such an argument, the debater would describe several representative examples of student athletes, then would show how each of them turned out to be successful in business. The point of the argument is not to merely describe the members of the group—the examples of student athletes—but to argue that the entire population of student athletes (or at least a substantial portion of them) shares the characteristic of becoming successful business people. Thus, the argument is designed to link examples chosen from a sample group of student athletes to the entire population of student athletes.
1705136452
1705136453
The above example illustrates a descriptive claim, the most common type of claim supported by argument by generalization. The argument only shows that they become successful business people. It does not explicitly state that being a successful businessperson is good or bad. This description however, like most descriptions, has evaluative dimensions. If one believes that successful business people are fundamentally unhappy, the argument could be used to argue that participation in athletics is a bad idea. If, on the other hand, one believes that becoming a successful businessperson is valuable—because it provides a good income, stability for a family, etc.—then this argument can be used to argue that participation in student athletic programs is valuable. Probably, most people believe that success in business is valuable, but the point is that, although the argument is explicitly and primarily descriptive, it contains implicit evaluative dimensions, as well.
1705136454
1705136455
Some arguments using generalization links are better than others. The adequacy of a generalization link is based on at least two assumptions: 1) that a sufficient number of examples are presented as evidence, and 2) that the examples are representative of the entire group. For example, to argue that student athletes become successful business people by pointing to only a couple of examples would not be sufficient as a generalization. In that particular case, two examples are not sufficient to allow the debater to make a statement about student athletes in general. The second question to be asked about the adequacy of a generalization link is whether the examples are representative of the group as a whole. If the examples of student athletes came, for instance, from a single university, one might not be able to argue that they are representative of the entire population of student athletes. More will be said about the adequacy of this and other kinds of links in Chapter 21, “Fallacies in Argumentation.”
1705136456
1705136457
Argument by generalization functions because several examples drawn from a group are linked to the overall group. This link allows debaters to create descriptive arguments. Other kinds of links can also be used to create descriptive and other kinds of arguments.
1705136458
1705136460
17.3 Argument by Analogy
1705136461
1705136462
In an argument by generalization, a claim about a group is based on information about selected members of that group. The link created by analogy is different. Analogy, based on an association of similarity, occurs when the arguer makes a claim about one member of a group based on the features of some other member. As a generalization link moves from specific cases to a generality, a link by an analogy moves from one specific case to another.
1705136463
1705136464
Two subtly different kinds of analogy will be discussed in this chapter, using the first kind, the debater argues that one example is similar to another; in the second kind, the debater argues that two examples are so similar in known regards that they should be expected to also be similar in unknown regards.
1705136465
1705136466
With regard to the first kind of analogy, a debater might simply want to establish the similarity between two examples. For instance, consider the claim that “Life in 21st century China will be like life in 20th century United States.” That argument makes a statement about 21st century China based on the similarities between it and the life in the United States in the 20th century. To make such an argument, an arguer needs to describe some features of life in 20th century United States, then show that those features are likely to be present in 21st century China. The similarities of the features of those two examples then allow the arguer to make the general claim that, “Life in 21st century China will be like life in 20th century US.”
1705136467
1705136468
The conclusion of that analogy is a general one about similarities between the United States and China. That type of analogy might be used for evaluative purposes. If life in 20th century US was good, life in 21st century China might be expected to also be good. So, in a general way, something we know about the first example (life is good in 20th century US) predicts something we do not know about the second (life will be good in 21st Century China). However, that prediction is only implied and is not an explicit part of the analogy. The prediction becomes explicit in the second kind of analogy.
1705136469
1705136470
In the second kind of analogy, a debater might want to argue that two examples are so similar in known regards that they are also expected to be similar in unknown regards. That kind of argument by analogy also uses two parallel cases. The cases are said to be parallel because they both contain known similar features. However, the first case contains a known feature that is unknown in the second case. This argument by analogy infers that the features known in the first case and unknown in the second probably are present in both cases.
1705136471
1705136472
Using the earlier example of China and the US, the arguer might focus on similarities between the two parallel cases, China and the US. The arguer might suggest that those two cases are similar in two known regards: the presence of strong economic power and increasing numbers of women in the workplace. Based on the presence of those two known similarities, the arguer then might infer that a known feature of the first case, e.g., a general lowering of the rate of unemployment, will also be present in the second case. Thus, the claim made in that argument could be that 21st century China will witness a general lowering of the rate of unemployment. The analogy allows the debater to infer that something known in the first case is present, although unknown, in the second case.
1705136473
1705136474
In many instances, an argument by analogy supports a descriptive argument. Consider again the general analogy that 21st century China will be like 20th century US. A precise reading of that claim suggests that it merely describes, without evaluating, life in 20th century US, and 21st century China. However, this and most analogies are at least implicitly evaluative because the audience evaluates the first case (here, life in 20th century US.) in a particular way and, therefore, will likely come to evaluate the second case (here, life in 21st century China) in the same way. Depending on the audience, people may have either a negative or positive association with 20th century US. They may see the US, as an arrogant superpower and a terrible polluter, or, on the positive side, may see the US as an economic powerhouse and an advocate for human rights. So, depending on whether the audience sees 20th century US as positive or negative, the argument by analogy will lead them to evaluate 21st century China in the same positive or negative way. Thus, an analogy has the function of transferring the positive or negative evaluation of the 20th century US to 21st century China.
1705136475
1705136476
A precise reading of the second example’s claim that “21st century China will witness a general lowering of the rate of unemployment” also indicates that this is a descriptive claim. A more insightful reading, however, suggests that the claim is also evaluative given the assumption that most audiences perceive lowering the rate of unemployment as a positive thing. Thus, a more subtle reading of this claim indicates that it is evaluative as well as descriptive.
1705136477
1705136479
17.4 Argument by Causality
1705136480
1705136481
Causal links, as the name implies, are used to construct arguments of cause and effect. Constructing an argument about cause and effect is especially difficult because causes cannot be observed; we can only infer them. Over the years, people have developed a variety of ways to infer cause and effect relationships. Four of those ways include absence and presence, change over time, correlation, and controlled empirical studies.
1705136482
1705136483
One common way to support a cause and effect relationship involves comparing the absence and presence of possible causes and effects to find potential relationships between the two. Using the idea of presence, a person might first notice the simultaneous presence of a suspected cause and suspected effect. Observers might notice that the US, a country in which citizens own a large number of guns, has a large number of murders—about 160,000 murders per year in a population of 300 million (Harris, 2006). Thus, the simultaneous presence of a suspected cause (a large number of citizen-owned guns) and a suspected effect (a high murder rate) might lead an arguer to suggest that the number of guns contributes to the high murder rate. But the simultaneous presence of cause and effect does not provide particularly good evidence of a cause and effect relationship. Someone might point to other factors in the US that could be the real cause of the high incidence of murder—factors such as racial strife or inadequate numbers of police in some major cities. In other words, alternative factors unrelated to the suspected cause (citizen-owned guns) might be the actual cause of the effect (high murder rate).
1705136484
1705136485
In order to argue that alternative factors such as those mentioned above are not the real cause, the arguer must look not only to the simultaneous presence of the suspected cause and effect, but to simultaneous absence, as well. In the above example, the arguer might look for an instance where the suspected cause (citizen-owned guns) is absent, pointing perhaps to the case of China where citizen-owned guns are illegal. The question then becomes whether the suspected effect (a high incidence of murder) is or is not present in China. The arguer could then point to the fact that the likelihood of being murdered in the US is four times higher than the chance of being murdered in China (Harris, 2006). Therefore, the arguer could conclude that because both the suspected cause(citizen-owned guns) and the suspected effect (high rate of murder) are absent in China, citizen gun ownership contributes to the high rate of murder in the US.
1705136486
1705136487
So, one method of supporting a causal relationship is to use the simultaneous presence and absence of the cause and effect. By itself, the simultaneous presence of a suspected cause and effect does not provide particularly strong evidence of a causal relationship. However, the added evidence of the simultaneous absence of the suspected causes and effects makes much better evidence for a cause and effect relationship.
1705136488
1705136489
A second way of arguing for a cause and effect relationship is change over time. When someone argues that one thing causes another, he or she frequently does so by observing that some change in the first thing is accompanied by a corresponding change in the second. The person is able to observe the changes in both, and, on the basis of those changes, can infer that the change in the first thing caused the change in the second. For instance, a man who changes from a high fat to a low fat diet and loses five kilos in a month might lead to his inference that reducing the fat content in his diet caused his weight loss. Two things were observed: a change in diet and a loss of weight. From those observations, the arguer inferred that the changes in diet caused weight loss.
1705136490
1705136491
A third way to make inferences about cause and effect relationships is to examine correlation. To use the method of correlation to infer a cause and effect relationship, an arguer can point to the fact that the incidence of a suspected cause and its suspected effect rise and fall in relationship to one another. For instance, the arguer will suggest that as the frequency of a suspected cause increases, the frequency of the suspected effect also increases. Of course, many are quick to state that correlation and causation are not the same things. Although correlation is not causation, correlation is one test of a causal relationship. If two events are not correlated, any inference that they are causally related to one another would be wrong. So, using the method of correlation, a debater might argue that smoking is one of the causes of lung cancer because, as the incidence of smoking increases within a society, the incidence of lung cancer increases, as well. Such a correlation provides evidence, although not perfect evidence, that smoking causes lung cancer. Correlation by itself may not be sufficient to prove a causal relationship, because correlation does not rule out other potential causes (such as living in a polluted area, being genetically disposed to lung cancer, etc.). Still, correlation is a method to make at least an initial inference regarding a causal relationship.
1705136492
1705136493
A fourth method of supporting a cause and effect relationship involves controlled empirical studies. As stated earlier, correlation and simultaneous presence of suspected causes and effects are imperfect methods of inferring causal relationships. Sometimes, scholars are able to design controlled empirical studies that help to offset those imperfections. For instance, an empirical study might examine the relationship between smoking and lung cancer while controlling for the effects of other possible causes (such as pollution, genetics, etc.) and might discover that, even considering the potential effects of those alternative causes, the causal relationship between smoking and lung cancer still is substantial.
1705136494
1705136495
Debaters frequently use argument by causality to judge actions based on their consequences. A debater first uses causal links to convince the audience that a particular action will cause certain consequences. In that case, the argument is that an action (smoking) leads to a consequence (lung cancer). Then, the debater can argue that an action that has good consequences is justified while one that has bad consequences is not. Thus, in the smoking example, the debater maintains, either implicitly or explicitly, that society has an obligation to stop people from smoking to reduce the negative consequences of that action.
[
上一页 ]
[ :1.705136446e+09 ]
[
下一页 ]