There are two steps to dealing with any argument:  1.  Identifying the type of argument, its premises and conclusion, and 2.  Evaluating the argument using the techniques appropriate for that type of argument.

First, a word about recognizing arguments.  Just because two statements are presented together, that doesn’t mean that it is an argument.  An argument requires more than just two statements; it also requires an implicit or explicit claim that one statement follows from the other statement, or that one statement is a reason to believe another statement.  This claim is often presented using the term “Therefore”, but not always.  We call this term and others like it a conclusion indicator word, since it is generally followed by a conclusion.  You should be aware of several conclusion indicator words, as well as several premise indicator words.

In this course we will consider three argument types:  Deductive and Inductive.

Deductive arguments are those which are intended to give an absolute guarantee that the conclusion follows from the premise.  I prefer to say that they are those arguments where the premises have a 100% relevance to the conclusion, meaning that the premises given are all that needs to be considered to judge the argument.  Examples:  Math, arguments based on definitions, syllogisms (hypothetical, categorical, and disjunctive)

Evaluating Deductive arguments:  There are always two steps to judging an argument: 1) Evaluating the inference, and 2) Evaluating the premises.  To evaluate a deductive inference, ask yourself this question:  If it so happens that the premises are true, would there be any chance that the conclusion is false?  If the answer is NO, then the inference of the argument is good, and the argument is called VALID.  Otherwise, it is invalid.

Next, if the premises are indeed true (and the argument is valid) the argument is called SOUND.

Inductive (and Abductive) arguments:  In both these types of arguments, the premises are relevant, but other information could alter the strength of the argument.

In many inductive arguments, we take a property of a sample, and apply that property to another group.  If the sample is included in the group, it is a generalization.  If the sample is not included in the target group, the argument is an analogy.  If the target involves some future event, the argument is a prediction.

To judge inductive arguments, we must consider the size of the sample, and whether it is representative.  If both of these conditions are reasonable, we call the argument strong, otherwise it is weak.

Then, if the premises are true (and the argument is strong), we call the argument cogent or compelling.

In some other inductive arguments (often called abductive arguments), we take certain observations, and we consider certain explanatory hypotheses.  We favor the explanatory hypothesis which makes the observations most likely.  If there is a strong likelihood based upon one hypothesis, and no other hypothesis makes the observations very likely, then the argument is strong, otherwise it is weak.

One way to evaluate these inductive arguments is called the Surprise Principle by Elliott Sober.  Assume that the hypothesis is true (only for the purpose of evaluation), how surprised would you be by the observations.  The more surprised you would be, the weaker the argument.

Then, if the premises are true, meaning that the observations stated were actually observed (and the argument is strong), we could call the argument cogent or compelling.  (You should know that often abductive arguments are not presented in this format.  Many times one simply states the observations, and then states that one particular hypothesis is the most likely explanation.)