Two separate methods are used to describe the world—categories and patterns—the first verbal, the second not. Verbal leads to logic and patterns lead to similarity, different methods of reasoning which proceed simultaneously.
A category is the basis of a word-oriented description of our world. It is produced and used of the language dominant side of the brain. When we think about the world, we think of sky, trees, people arriving, departing, asking for, and delivering to. We describe our situations to our friends and they do the same for us. Alone and struggling to achieve a personal goal, we might list the factors that could lead to success or that might cause problems.
It must be noted that using categories to describe the world invariably entails a loss of specific detail. If we describe our backyard is full of trees, the details of species are gone.
Another consequence of using words is that only those features described by words are then available in the contemplation of our world.
A pattern is a non-verbal method of describing the world. It is produced and used in the non-language brain hemisphere. It’s easiest to think of a pattern as a portion of a visual snapshot of the world, but patterns also exist in sounds that we hear but don’t name, taste that we crave but can’t name, and so on.
Patterns can capture more details than words. For example, a new mustache on your pal. His name remains the same, but his pattern holds the new feature.
Your friend’s mustache can be added to your verbal description of the world. That is certainly true. It can be done in two ways—as a tag-on description of your friend or a part of the meaning of his name.
Since the brain is so interconnected, categories are compared with patterns, repeatedly though their progress up the concept elevator from sense data to integrated world view to decision-making and back down to speech or action. This comparison can result in the pattern conforming to the category or the category being supplemented with additional description.
When the pattern conforms to a category, any additional detail in the pattern is discarded for its subsequent uses. You might wonder how could there be additional detail in the pattern since it conformed to the category. That occurs because the conformance test is whether the neural threshold has been reached, which does not require an exact match.
Note: since the capacity of working memory is limited, the additional verbal description expands the space required to process the category, restricting the number of categories that can coexist in working memory.
Logical deduction is the method of reasoning with categories. This is word-based logic.
Aristotle’s syllogism is the classic example: All men are mortal. Socrates is a man. Therefore, Socrates is mortal. Deductive logic has progressed far beyond that simple example, but deduction remains word-based.
Logical deduction is often called reasoning from the general to the specific. Logical deduction’s conclusions are guaranteed exactly true, when all its starting facts are true.
Induction is a method of reasoning based on patterns. It uses similarity between patterns, non-verbal organizations of features, as a basis for treating patterns as the same or different.
A typical induction occurs when one compares the present situation with a past situation and finding enough similarities decide that the outcome of the current situation is likely to be the same as the outcome of the past situation. The action of the stock market is one such case.
Induction is often handily described as reasoning from the specific to the general.
How close the similarity must be – how many features in common must exist for a person to have assurance in their conclusion? That question highlights an essential problem with induction. Its conclusion is not guaranteed to be true, even when its starting facts are true.
Why Ever Use Induction?
If deductive conclusions are always valid, why would anyone ever use induction?
Three basic reasons
- We don’t always have time to reach the logical conclusion before we need to act.
- The raw information does not always fit into neat categories.
- The conclusions we search for do not always logically follow from the information we have.
In those cases, since deduction won’t serve, we use induction. It may be flawed, but it does more quickly and easily.