We all do it. We pounce on an offending statement in a discussion by saying it isn’t logical, as if this alone should settle the matter. With no less conviction, we defend our own arguments by saying that they are. This vernacular tendency seems to hold out logic as an arbiter of truth, or at least as a way of getting closer to it.
This may be especially the case for the “rules” of deductive logic. They seem so ironclad when laid out as syllogisms, and accepting the premises makes the conclusions seem so obvious. Consider the following example of a type of deductive reasoning called implication. If you buy the first two propositions, the conclusion is hard to resist.
If there is combustion, then oxygen must be present.
There is combustion.
Oxygen is present.
A special kind of verbal behavior?
As behavior analysts, we recognize that what is commonly called logic is verbal behavior. It’s not so obvious how we might distinguish logic from other aspects of verbal behavior, however, or even if we should. Is it a special kind of verbal behavior? Are there even different “kinds” of verbal behavior?
Chapter 6 addresses this issue by analyzing reasoning – a more general reference to logic – in terms of the fundamental verbal response classes Skinner introduced in his seminal book, Verbal behavior (1957). They come together in the form of propositions – statements about what we know, believe, think, judge, or assume (e.g., The sky is blue.) These statements involve tacts (flower), assertive autoclitics (The flower is yellow.), quantifying autoclitics (reference to some flowers or all flowers or no flowers), and more. When we emit propositions that may be readily accepted by a listener, logicians call them premises, and the proposition that is more likely to be accepted as a result is called a conclusion. Yes, it can get a little complicated, so the chapter requires more than a quick read.
What you will learn is that there is nothing special about the logicality of verbal behavior. Verbal behavior we loosely describe as reasoning can indeed be impressive, but it’s still “just verbal behavior” (see Chapter 2) that can be decomposed into its component response classes. These exercises are admittedly complex and laborious, but it’s important to avoid any slippage in translating colloquial views of logic into what our science has revealed about verbal behavior.
If it’s just verbal behavior…
Verbal behavior that can be described by the rules of deductive reasoning may be correct or true in nature, but that’s not because it’s consistent with these rules. Similarly, statements that violate deductive rules may be untrue, but it’s not their illogicality that makes them untrue. Our verbal behavior does not control physical circumstances, aside from influencing the behavior of others. In other words, the validity of deductive reasoning is not a criterion for the truth of statements about events. Although deductively valid reasoning may often correspond with actual events, it does not have to and sometimes does not. For example, it may be valid reasoning but based on a false premise.
What this means is that accusations of illogicality as a way of criticizing someone’s argument misunderstand the nature of logic as verbal behavior. What we call the rules of deductive logic are only regularities in how propositions are stated. These regularities are the basis for reinforcement practices in verbal communities that include instructors who teach courses in logic, philosophy graduate students, and hardly anyone else. The result is called the rules of deductive reasoning. The rules of grammar come from the same kind of reinforcement practices involving grammarians and English teachers.
Denigrating an argument because it is illogical suggests that the speaker buys into the point of view that logicality has a special status that goes beyond ordinary verbal behavior. From this everyday point of view, reasoning, if not all verbal behavior, is a symbolic activity that involves private subjective meanings that exist in a logical mental domain. This perspective involves misunderstandings that go well beyond confusions about the nature of reasoning (again, see Chapter 2).
To return to the title, logic doesn’t matter because it is based on fundamental confusions about the nature of verbal behavior. The logicality of an argument is not a meaningful basis for assessing its correctness because the notion of logicality involves a mentalistic view of verbal behavior that is at odds with our scientific understanding of operant learning and verbal behavior. If you are tempted to demolish your friend’s argument by saying it is illogical, look instead for other reasons it may be wrong. It may well be, but bad logic isn’t the reason. And remember: verbal behavior is not fundamentally different from non-verbal behavior – it’s all learned the same way.
Skinner, B. F. (1957). Verbal behavior. New York: Appleton-Century-Crofts.