Lessons from a French Washing Machine
(Note: This commentary was written during Dr. Lattal’s sabbatical year in France)
My apartment in France came with a washer/dryer combo unit, which is simply wonderful. The last time I lived in here, my apartment was sans lave-linge. This meant that I had to use the laundromat down the street, which always was an adventure and occasionally a not-pleasant one. Today I happily toted my first load of laundry to my washing machine, only to encounter a dazzling array of bells, whistles, and buttons on its control panel that I could not decipher. Two choices confronted me: I could fidget with the gadgets and through trial and error make it work (the buttons were not even labeled with words, only icons that were unfamiliar to me) or I could get some instructions. A manual for the machine was not to be found in the apartment, so I called on my English-speaking neighbor, who I knew had a similar machine, for help. Within minutes of his arrival, my dirty clothes were happily swimming around in a broth of laundry soap.
Psychologists label trial-and-error learning as contingency-shaped behavior and learning through rules or instructions as rule-governed behavior. Both have their value in establishing new skills. Trial and error is valuable because we learn not only what to do, but also what doesn’t work, or what doesn’t work as effectively. It often is said that there is no substitute for experience, and experience is what we gain with trial-and-error learning. Rule-governed learning is valuable because it saves time – we benefit from the trial-and-error experiences of others.
Both also have drawbacks. As with my encounter with the lave-linge, learning to operate it via trial-and-error learning would have been a big waste of time, time that I could have used more effectively in doing other things. Because alternatives were not tried, rule-governed behavior can be somewhat rigid and inflexible. Not knowing whether variations in the solution also will work, because they have not been experienced through the trial-and-error process, makes trying new things less likely. “A solution” sometimes is seen as “the solution” when rules govern the behavior.
A difference between contingency shaped and rule governed behavior can appear when the current solution no longer works, that is, when the previously reinforced behavior no longer solves the problem. Resurgence refers to the tendency of previously established behavior to recur when a current response is no longer reinforced (i.e., extinguished). If a solution has been established through trial and error, there is a large collection of previous behavior that can resurge or recur, and that behavior can serve as the basis for a new solution. On the other hand, a solution based on a rule has come about in the absence of establishing a history of other behavior on the way to a solution. When a solution fails, there is no previous history of responses to resurge, so the development of a new solution may be slower to occur.
In practice, both contingency shaping and rule governance have their place in teaching people new skills. Some problems require quick solutions that don’t waste time, while others are best solved by giving the solver practice that may prove useful in solving new, related problems in the future. Some solutions once emplaced don’t require much further tinkering, my lave-linge being an example. Other solutions are more interim and warrant continuing experimentation. Some solutions are a mix of the two. A contingency shaped solution often becomes a rule for subsequent learners and rule-governed solutions get modified through experience.