David Stubblefield, Senior Editor

The famous 18th century, Scottish philosopher David Hume claimed that we cannot know for certainty that the sun will rise tomorrow. His reasoning rested on his larger belief that we cannot assume the uniformity of nature or, in his words, that “instances of which we have had no experience resemble those of which we have had experience” (Hume 89).

Bertrand Russell, a renowned English philosopher, writing in the 20th century, extended Hume’s concerned and argued that even barnyard animals would benefit from a healthy dose of Hume’s skepticism. In fact, Russell insisted that failing to do so could be a matter of life and death. To do this, he asserted the likelihood that upon the appearance of his owner and his standard feeding time, the chicken would infer that he was likely to be fed. However, Russell insisted that this assumption could always be erroneous because it is likely that one day “the man who has fed the chicken every day throughout its life at last wrings its neck instead, showing that more refined views as to the uniformity of nature would have been useful to the chicken” (Russell 123)

Establishing Regularities in our Experience

While Hume and Russell show how difficult it is for human reason to attain absolute certainty when reasoning from experience, their comments also throw into relief the necessity of making such inferences and assuming that there are regularities in the world. Indeed the chicken has to make this assumption if he is going to get his fair share of food. Likewise, we cannot know with absolute certainty that our 8:00 am class will continue to meet at 8:00 am, but in the absence of any contradictory evidence, we have to assume that it will if we do not want to be marked absent. In fact, if we doubted all regularities of experience, then it would be impossible to accomplish anything or be effective in the world. In other words, we would be unable to base our decisions about what route to take to class or where to eat today on anything but pure whim.

The point is that we find ourselves in a very similar position to Russell’s chicken. We all make generalizations from our experience every day. And while nature may not be uniform, leaving open the possibility that tomorrow will not resemble today, we can still make reasonable inferences from particular instances we have observed. When we do so, we are making what logicians call inductive arguments.

Analyzing Inductive Arguments

Induction is a species of reasoning that involves making probable generalizations about the future from particular empirical instances or facts from the past. This move from particular to general is also commonly described in rhetoric textbooks as an inference from sample to population.

Within this context, it makes sense to speak of the relative strength of a given inductive argument. For example, consider the following cases:

Example #1: My roommate never studies and has a 4.0 GPA; therefore, studying is not a necessary part of being a successful student.

Example #2: In a survey of one hundred American universities, it was found that students with a 4.0 or higher studied an average amount of 30 hours a week; therefore, putting in lots of hours studying is essential for being a successful student.

In the first case, while the premise may be true—the author’s roommate may, in fact, be excelling in school without studying—the conclusion is not warranted since this is a very limited sample.
In the second case, we can see that a much more extensive sample was taken, making the conclusion much more probable; however, once again, the conclusion is not certain since it is always possible that there is a genius that doesn’t need to study.

Problems with the Numbers

While sample size often provides some indication of a good inductive argument, things are not always that simple. Sometimes a particularly significant example can be very persuasive for a specific audience. For example, from the fact that one person was able to hack into a computer system, a group of computer engineers might infer that the computer’s security system is not acceptable. Likewise, within some communities, a single expert’s testimony might be very convincing, as when a world-renowned scholar shares her opinion on a certain matter. In other words, though sample size is certainly something to consider, there are many cases where sample size fails to fully capture the persuasive capacity of an inductive argument.

Yet when our sample size is too small, our inferences run the risk of becoming problematic. We may have had our instructors write something like “generalization” on our papers when the inference from particular evidence to general conclusion does not seem warranted. In real life, we are likely to call such hasty generalizations stereotyping. For example, we may reason inductively that since the best students we have known in the Mechanical Engineering department are males, women do not make good mechanical engineers. Of course such an erroneous conclusion is not only based on a small sample size but may, in fact, neglect other factors such as the social emphasis on femininity, discrimination in STEM fields, our evaluative process to determine “best,” and noteworthy exceptions to our conclusion.

What Makes an Inductive Argument Good?

Thus, we can see that inductive arguments cannot only be evaluated in terms of their reasonableness but also in terms of their ethicality. In the light of such potential for bad or unethical uses of induction, you might ask what constitutes good, effective, and ethical uses of inductive reasoning?

While there is an answer to this question, it is neither simple nor steadfast; rather, like most things in rhetoric, it is complex and contingent. According to The Stanford Dictionary of Philosophy, “There is no comprehensive theory of sound induction, no set of agreed upon rules that license good or sound inductive inference, nor is there a serious prospect of such a theory.” In the absence of any consensus on this matter, we can think of making good inductive arguments as a situated art form, one which requires a good bit of familiarity with non-logical factors or social factors such as cultural and political norms, how various forms of authority function in different contexts, and the kinds of genre norms that typically govern certain types of symbolic action. And while there is no definitive method that can guarantee good inductive arguments, as we analyze and participate in specific rhetorical situations, we learn more about what constitutes a good inductive argument within those situations.

(1). Hume, David, [THN] 1888, Hume’s Treatise of Human Nature, edited by L. A. Selby Bigge, Oxford, Clarendon Press. Originally published 1739–40

(2). Russell, Bertrand. “Induction” in The Basic Writings of Bertrand Russell: 1903-1959. Egner, Robert E., and Lester E. Denonn, eds. Simon & Schuster, 1967