Edward Thorndike - Behavioral Psychology - The Canon

Psychology: Essential Thinkers, Classic Theories, and How They Inform Your World - Andrea Bonior 2016

Edward Thorndike
Behavioral Psychology
The Canon


BORN 1874, Williamsburg, Massachusetts


DIED 1949, Montrose, New York


Educated at Wesleyan University, Harvard University, and Columbia University



Edward Thorndike’s law of effect was his biggest idea. It is behaviorism in a nutshell. It states that a behavior that evokes a pleasurable response is more likely to be repeated than a behavior that evokes an unpleasant one. This became a fundamental tenet of later conditioning theories. Thorndike eventually refined this idea even further, saying that rewards work much more effectively to strengthen learned responses than punishments do to deter undesired responses.

Thorndike also put forth the law of use and the law of disuse. The former says that the more frequently a stimulus and a response are connected, the more ingrained the resulting association will become. And the latter is its counterpart—the longer a stimulus and a response have been disconnected, the less ingrained the connection will become. Thorndike built on Pavlov’s theories of classical conditioning by delving further into the concept of learning; in fact, we can consider him the first educational psychologist. Like the behaviorist pioneers before him, Thorndike was an empiricist, interested above all in measuring and quantifying his findings.

Thorndike called his overarching theory connectionism. This term refers to the application of conditioning principles to learning; the name came from the idea that learning is always the result of the connection between a stimulus and a response. Thorndike also did work that altered the way reading and spelling were taught. He analyzed the frequency of words found in various children’s works, and from them he compiled dictionaries specifically for children. His rationale was that language learning would be improved through greater focus on common words than on less common ones. That no one before Thorndike had tried such a rational, practical approach may seem surprising in retrospect, but it took Thorndike’s interest in the intersection of data and education to bring that change about.


Thorndike created scales to assess students’ growth across various areas, from arithmetic to handwriting, and these led to his development of several intelligence tests. And he advanced the idea that intelligence has different aspects—that a person can be strong in one area but not in another, or that someone can understand advanced calculus but not have the spatial wherewithal to figure out which way to turn in order to get back on the freeway.

Also credited to Thorndike’s research is the concept of trial and error, which he developed by studying cats and their process of learning how to escape from an enclosure, called a puzzle box, by pressing a foot pedal. As the cats tried to escape, they made errors, but they made fewer and fewer errors as the experiment went on. Thorndike’s key finding was that multiple trials and multiple errors are necessary components of the learning process.

Tasks in which participants exhibit the pattern of a learning curve are those in which participants display gradual growth and decreased mistakes, but mistakes are not eliminated all at once. If sudden insight were to eliminate all errors, then the typical learning curve would show a steep spike in competence, but in fact the typical learning curve shows sloping upward progress. This is why, in our modern-day use of the phrase learning curve, we are getting it 100 percent wrong. A steep learning curve, which many people associate with something that will be hard at first to learn, actually indicates quick and easy improvement in performance. It’s important to remember that the curve represents improvement in performance, not the difficulty of the task.

KEY EXPERIMENT Edward Thorndike used animal research to formulate many of his additional theories on learning, and his version of the puzzle box was instrumental in many of his findings. It had wooden slats that the animals inside could see out of, and it was rigged to offer a particular way of escape—often a foot pedal, as we saw in his work with cats. (In a departure from other puzzle boxes that had been developed, this means of escape was inside the box itself, and so it could be used to test a variety of animals, not just monkeys, which could use their hands to reach outside the box; in fact, Thorndike made various alterations to the box so he could test chickens, dogs, and even fish.)

In a typical experiment, a cat, annoyed at being trapped in the box, would at first sulk around and make its unhappiness known, in that unmistakable though not particularly endearing way that cats have. This reaction was particularly exacerbated when food was placed outside the box, in the cat’s direct view. At some point, the cat would press the foot pedal by accident, an action that led to the happy surprise of the door’s springing open. Of course, the cats got faster and faster with subsequent practice, though eventually they reached a maximum speed. They couldn’t get better indefinitely, and their performance would finally plateau. Even if overall speed differed from animal to animal, the pattern of how they improved was similar. As we know, Thorndike referred to the plotted pattern of this improved performance as a learning curve, and he showed its relevance to a variety of different animals, including humans.


Thorndike’s work created the foundation for the study of learning and for the field of educational psychology as a whole. His theories also led directly to further developments within conditioning theory and were a direct influence on B. F. Skinner, whose theories of operant conditioning were an outgrowth of connectionism.


Have you ever looked at your child’s report card and wondered just what methods were used to quantify her learning? Though controversy over standardized testing has come to a fever pitch, Thorndike’s idea that learning can be assessed and measured, just like any other behavior, and that different teaching techniques can be pitted against each other, to determine which one will lead to better outcomes, was a revolutionary notion that still underlies education standards today. And, when you really think about it, isn’t it a relief that there are at least some specific standards by which your child can be assessed? If your child’s teacher weren’t able to point to concrete, quantifiable metrics for math tests or book reports as the justification for an A, a B, a C, or a D, then what would the grade really mean? At best, it might not have much value; at worst, it might be biased.

Perhaps you take it for granted that practice makes perfect, but have you ever really thought about the mechanics of just why that’s the case? It’s because of trial and error, not mere repetition; the difference between the two is something Thorndike discovered and emphasized in his theories. Let’s say you’re trying to master a complex piano piece. You know that you’ll get better the more you play it, but the reason is that with each trial, you are making errors—some big, some small—and with each of those mistakes, you take in information about exactly how you made the miscalculation and what you need to do to improve.

Let’s think further about how you learn. Perhaps you’ve turned to the Internet to master a particular skill—making a hundred origami swans for your sister’s wedding, installing your own garbage disposal, or making a perfect omelet. You may turn to YouTube, which has revolutionized the world of how-to instruction with millions of step-by-step videos. But what is the most helpful way to use the video—to watch it through once, close your laptop, and then go on your merry way, or to keep the video alongside you as you learn, stopping and starting as you try to master the task on your own? You may have noticed that you fare much better the second way, even if you thought you had it after watching the first time. Why? Any child can tell you: We learn by doing, not just by watching. You will learn to perform a certain procedure or behavior much better if you can actually participate in the process and can keep refining your technique through trial and error rather than just watching someone else do it, which keeps you from going through the learning curve yourself.

Thorndike’s notion of connectionism also underlies motivation for learning. Recall his theory that the connection between the stimulus and the response must remain strong, and that a behavior that creates a pleasurable response is more likely to be repeated than a behavior that creates an unpleasant response. Now think of trying to help your child learn multiplication tables. When she finally gets an answer right, do you clap and praise and give a high five? Or do you say, “Okay—finally”? That would be the way to create a negative consequence for learning, whereas connectionism would tell us that with praise, the right answers will start to come much more frequently. Anyone who’s learned that a preschooler responds much better to praise for doing what’s right than to scolding for not having completed the same task sooner has seen Thorndike’s theories in action.