1 HANDBOOK (New or substantially modified material appears in boxes.) I. ARGUMENT RECOGNITION Important Concepts An argument is a unit of reasoning that attempts to prove that a certain idea is true by citing other ideas as evidence. The idea that the argument tries to prove is called the ultimate conclusion. Ideas that the argument uses as evidence for the ultimate conclusion, but that the argument assumes to be true without providing proof, are called premises. Intermediate ideas on the way from the premises to the ultimate conclusion are called subconclusions. The connection that holds between a set of ideas, R, and another idea, C, when the truth of the ideas in R is supposed to establish the truth of C is called an inference. In order to identify an argument, we ask, Is this passage trying to convince us that something is true? If so, the passage contains an argument. If not, it doesn t. Use Inference Indicator Expressions We can sometimes recognize an argument by spotting inference indicator expressions, although we should remember that some passages containing inference indicator expressions don t contain arguments because inference indicator expressions can occur in explanations as well. We should remember that some passages containing arguments don t contain inference indicator expressions. Reason indicator expressions show that X is being given as a reason to believe Y. Some examples: Y because X or Because X, Y. Y, since X or Since X, Y. Given that X, Y or Y, given that X. Assuming that X, Y or Y, assuming that X. Inasmuch as X, Y or Y, inasmuch as X. In view of the fact that X, Y or Y, in view of the fact that X. Y. The reason is that X Y. After all, X. Conclusion indicator expressions show that Y is supposed to be concluded from X. Some examples: X. Therefore Y. X. Thus Y. X. Consequently Y. X. Hence Y. X. So Y. X. This goes to show that Y. X. It follows that Y. X. As a result, Y. X. That s why Y. X, which implies that Y. X, which means that Y. To determine whether an inference indicator expression is a reason indicator expression or a conclusion indicator expression, first replace it with because, then replace it with therefore, and see which passage is more like the original. II. ARGUMENT ANALYSIS Identifying the important ideas, identifying the argumentative role of the ideas, identifying the inferences, and reconstructing the argument are often intermingled in practice and are frequently done mentally.
2 1. Identify the important ideas We make a list of the important ideas in the argument. Ideas are complete thoughts that are either true or false, even though we might not know for certain which it is. Start With the Ultimate Conclusion We start by asking, What is the main idea that this argument is trying to get us to believe? This main idea is the ultimate conclusion and we write it first in our list of important ideas, giving it the number 1 and putting a U next to it. If we can t identify the ultimate conclusion, we should simply list the important ideas in the order they appear in the passage. Once we draw in the inferences, the ultimate conclusion will be the idea to which all of the other ideas eventually lead. Identify the Lines of Reasoning It s often useful to determine whether or not the argument appears to have more than one line of reasoning as soon as possible. If an argument does have more than one line of reasoning, it s helpful to count how many separate lines of reasoning the argument has and to determine which ideas belong to each line. This will substantially reduce the complexity of the diagramming process. Identify the Other Important Ideas After we ve identified the ultimate conclusion, we go back and record all of the other ideas that strike us as relevant to establishing the truth of the ultimate conclusion, and number them starting with 2. If we aren t sure whether or not an idea is important, we should include it just to be safe. It s okay if some of the ideas in our list aren t included in the eventual diagram. Sentences and Ideas Sentences that convey ideas are usually statements. To be complete, a statement must have a subject and a predicate and all sentence connectors must be connecting sentences that are themselves complete. Masking statements, unlike normal statements, convey ideas they don t actually state. If the idea is important, we should rephrase the statement and include it in our list of ideas. Unimportant statements aren t relevant to establishing the truth of the ultimate conclusion. We won t include them in our list of ideas. Normal questions don t convey ideas. Statement questions do convey ideas. If the idea is important, we should rephrase the question and include it in our list of ideas. Normal commands don t convey ideas. Statement commands do convey ideas. If the idea is important, we should rephrase the command and include it in our list of ideas. Sometimes a sentence contains more than one idea. In general, sentence S conveys idea I if the truth of S ensures the truth of I. We must divide a sentence into its component ideas around inference indicator expressions. We may divide a
3 compound sentence around connectives like and, but, yet, however, although, even though, moreover, and nevertheless. We should split a sentence into its component ideas if the component ideas significantly differ in their plausibility or if one of the component ideas appears without the others elsewhere in the argument. We can t divide a compound sentence around connectives like if then, or or. Sometimes multiple sentences in an argument convey the same idea. S1 and S2 convey the same idea just in case if S1 is true then S2 is true and if S2 is true then S1 is true. We write down each important idea only once, even when it s conveyed by more than one sentence. Use Inference Indicator Expressions Inference indicator expressions can help us decide if an idea is important. If an idea is the object of an inference indicator expression, then it s either the reason or the conclusion of an inference and so needs to be included in the argument. 2. Identify the argumentative role of these ideas If we know what the ultimate conclusion of the argument is, we put a U next to it. For each of the other ideas in our list, we ask, Does the argument give us reason to believe this, or does the argument just take it for granted? If the argument doesn t give us reasons to believe an idea, it s a premise. We put P s next to the premises. If the argument does give us reasons to believe an idea, it s a subconclusion. We put S s next to the subconclusions. If we don t know what the ultimate conclusion is, we can simply put C s (indicating generic conclusions) next to ideas that the argument gives us reason to believe. Use Inference Indicator Expressions We can use inference indicator expressions to help us. The ultimate conclusion can be the object of conclusion indicator expressions but not reason indicator expressions. Premises can be the object of reason indicator expressions but not conclusion indicator expressions. Subconclusions can be the object of reason indicator expressions and conclusion indicator expressions. Note Inference Eraser Expressions The connectors and (not and so ) and its equivalents, such as moreover, and the connector but, and its equivalents, such as yet, however, although, even though, nevertheless, tend to show us that an inference is not present between the ideas they connect. 3. Identify the inferences
4 We can focus on the conclusions (whether the ultimate conclusion or a subconclusion) and ask What reason does the argument give us to believe this? or we can focus on the reasons (whether a premise or a subconclusion) and ask What is the argument taking this to establish? We draw an arrow pointing from an idea to the idea that it s taken to support. I1 and I2 are dependent reasons in support of I3 if neither I1 nor I2 can support I3 alone but together they can support I3. We connect dependent reasons with a bracket and draw one arrow from the bracket to the conclusion of the inference. I1 and I2 are independent reasons in support of I3 if both I1 and I2 could support I3 alone. We draw separate arrows from independent reasons or lines of reasoning. Use Inference Indicator Expressions We can make use of inference indicator expressions to help us here, if the passage has them. Note Inference Eraser Expressions We should be alert for inference eraser expressions. Arrow In and Out Rules The ultimate conclusion must have at least one arrow pointing to it but no arrows pointing from it. Premises must have arrows going from them but no arrows going to them. Subconclusions must have arrows going to them and from them. Identify Dependent Reasons. There are seven tests for dependent reasons: 1) The Ophthalmology Test, 2) The Inference Indicator / Eraser Test, 3) The Try It Out Test, 4) The Puzzle Piece Test, 5) The Normative Conclusion Test, 6) The Comparative Conclusion Test, 7) The Means / Ends Test. Identify Independent Reasons To identify independent lines of reasoning, we ask How many separate lines of reasoning are we given? We can answer this question by identifying distinct themes that are advanced in support of the conclusion. These themes are roughly identical with the notions shared by different ideas. Instances of a generalization may be treated either as examples or as evidence. If they re treated as examples, they shouldn t be included in the diagram. If they re treated as evidence, they should be included in the diagram.
5 Double-Checking the Inferences We double check the inferences by reading away from the arrow head with a reason indicator expression, by reading toward the arrow head with a conclusion indicator expression, by reading the bracket as an and, and by comparing our inferences against the original argument. 4. Reconstruct the argument We reconstruct the argument by diagramming it. We refer to the ideas by number, put the number of the ultimate conclusion at the bottom, the numbers of the premises at the top, numbers of subconclusions in the middle, and use arrows to represent the inferences. We connect dependent reasons with a plus sign and we draw separate arrows from independent reasons. We label the arrows with capital letters to make them easier to refer to later. III. ARGUMENT EVALUATION 1. Appreciate the general structure of the argument A good argument establishes the truth of its ultimate conclusion and gives its audience good reason to think that the ultimate conclusion is true. A bad argument either doesn t establish the truth of its ultimate conclusion or else doesn t give its audience good reason to think that the ultimate conclusion is true. For arguments with only one line of reasoning, one bad premise or one bad inference is enough to make the argument bad. Arguments with independent lines of reasoning are good if even one of the lines of reasoning is good. The Hanging Man Model We can imagine that the ultimate conclusion of the argument is a fellow hanging onto one or more ropes (inference) suspended from one or more beams (premises). Each inference corresponds to a different segment of the rope, and each dependent premise corresponds to a different part of the beam. An argument is good if it holds the fellow up and bad if it lets the fellow fall. Evaluating Subconclusions We should never evaluate subconclusions as a part of the final evaluation of an argument. We may look at subconclusions in the process of evaluating an argument. If we disagree with a subconclusion, we should examine the premises and inferences above it. 2. Evaluate the premises
6 When evaluating a premise, we should ask ourselves three questions: 1) Is this premise true? 2) Would most members of the argument s audience, including people who don t already believe the ultimate conclusion, believe this premise? and 3) Does the argument s audience have good reason to believe this premise? If the answer to one of these questions is no, the premise is bad. If the answer to all three questions is yes, then the premise is good. Evaluating If then Sentences In order to evaluate an If then sentence, we ask ourselves Could the first part be true and the back part be false at the same time? If the answer is Yes, then the If then sentence is false. If the answer is No, then the If then sentence is true. 3. Evaluate the inferences To say that the inference between R and C is valid is to say that if R were true then C would have to be true as well. To say that the inference between R and C is good is to say that if R were true then C would most likely be true as well, although it wouldn t have to be true. To say that the inference between R and C is bad is to say that even if R were true, C could very easily be false; it s to say that the truth of R has virtually no bearing upon the truth of C. The Bob Method Bob is a perfectly gullible but perfectly rational person. We tell Bob to believe R and then ask ourselves In light of his belief in R, how likely is Bob to believe C? If Bob is compelled to believe C, then the inference between R and C is valid. If Bob is inclined but not compelled to believe C, then the inference between R and C is invalid but good. If Bob is not at all inclined to believe C, then the inference between R and C is invalid and bad. The Counterexample Method When evaluating argument A1, see if you can find a structurally similar argument, A2, that has true premises and a false conclusion. If you can find such an argument A2, then there s something wrong with at least one inference in A1. The Formal Method Determine if an inference has one of the following forms and evaluate it accordingly. If P then Q. P. Therefore Q is valid. If P then Q. Not Q. Therefore Not P is valid. If P then Q. Q. Therefore P is invalid. If P then Q. Not P. Therefore Not Q is invalid.
7 If P then Q and If Not Q the Not P, are logically equivalent and interchangeable. 4. Evaluate the argument We evaluate the argument in light of our evaluation of the premises and the inferences. Finding out that an argument is bad gives us no useful information about the ultimate conclusion because bad arguments can have true or false conclusions. Finding out that an argument is good does give us useful information about the ultimate conclusion because good arguments must have true conclusions. If we think that an argument is good, we should believe the ultimate conclusion. If we think that an argument is pretty good but not perfect, we should think that the conclusion is probably, but not definitely, true. If we re faced with arguments for competing positions, we should believe the position supported by the strongest arguments. Recognizing, Analyzing, and Evaluating Arguments in Real Life Unless an argument is particularly long, difficult, or important, we ll probably end up doing much of the analysis and evaluating mentally instead of on paper. We can do this by: identifying the ultimate conclusion of the argument, determining what other ideas are important, determining how these ideas relate to each other in the argument (e.g. where the inferences are, and so on), assessing the premises and inferences. IV. ARGUMENT CONSTRUCTION 1. Determine the ultimate conclusion We determine our ultimate conclusion by posing a question, considering various answers to the question, learning and thinking more about the issues involved, and formulating our answer to the question. The answer we settle on will be the ultimate conclusion of our argument. 2. Construct the chain of reasoning We construct our chain of reasoning by asking What are some reasons to think this idea is true? Once we have some ideas down, we diagram our argument by determining what argumentative role we intend each of each of these ideas to serve, deciding how we want our inferences to run, and then diagramming our argument.
8 We evaluate this chain of reasoning by first assessing the inferences. If an inference is weak, can we repair it by adding a dependent reason to plug the gap. We can use the tests for dependent reasons to find the dependent reason needed to strengthen an inference however, we should add the dependent reasons necessary to perfect an inference in our argument only if the original inference was sufficiently weak to justify the additional complexity involved in supplying the extra ideas. After we ve repaired our inferences, we assess our premises. If a premise isn t true, we change it to something that is. If a premise is true but might not be acceptable to the argument s audience, we make the premise a subconclusion by asking What are some reasons to think that this idea is true? and returning to the beginning. We then evaluate the new inferences, and new premises, repeating the process until our argument is good. In order to supply independent lines of reasoning for an argument of our own, we recognize the theme of the argument we ve already constructed and try to construct and argument of a completely different type to support the conclusion at hand. 3. Communicate the argument The passage containing our argument should be well written and easy to diagram. General Writing Advice Word Choice: We should use our working vocabulary, and we should avoid specialized terminology like ultimate conclusion, subconclusion, premises, or inference. Sentence Structure: We should use complete sentences; in particular, we should make sure that all of our simple sentences have a subject and predicate and that all of our connectives in our compound sentences are connecting smaller complete sentences. Sentence Variation: If we wish, we can use unimportant statements to set the stage for our argument, and express some important ideas as statement questions, statement commands and masking statements. We can add some normal question or command sentences for rhetorical flourish. Placement of the Ultimate Conclusion: We should generally put the ultimate conclusion near the beginning of the argument, unless it s controversial, in which case we should put it near the end of the argument. We may put our conclusion near the middle of a passage, as long as we put it between independent lines of reasoning. Placement of the Other Ideas: To make our argument as easy as possible to diagram, the proximity of the ideas in our passage should reflect the proximity of the
9 ideas in the diagram. We may repeat ideas, if this will help our reader to understand how these ideas work together. Highlighting Inferences: We should use inference indicator expressions to make the argument easier to understand whenever we think that an inference would be hard to recognize without them and we should write the argument vertically, up and down the arrows, rather than horizontally. Highlighting Independent Lines of Reasoning: If our argument has independent lines of reasoning, we should take care to help our readers to individuate them.