Tuesday, May 31, 2022

Frank C. Keil's "Wonder: Childhood and the Lifelong Love of Science"

Frank C. Keil is Charles C. & Dorathea S. Dilley Professor of Psychology at Yale University, where he is also a member of the Cognition and Development Lab. He is the author of Developmental Psychology: The Growth of Mind and Behavior and other books.

Keil applied the “Page 99 Test” to his new book, Wonder: Childhood and the Lifelong Love of Science, and reported the following:
From page 99:
… children, are wizards with computers in ways that I will never know. In a few minutes they are immersed in novel video games while I am still trying to move my agent. They figure out how to use new apps or software without reading any instructions and can seamlessly blend many software programs together to accomplish a task. I surrender completely to their greater abilities in this area. But what have they mastered? If today’s computers were cars, have they learned how they work or have they merely learned how to drive them? They seem to have become experts at learning user interfaces, not at how the underlying code or physical components work.

Perhaps twenty-first-century children are learning a higher level of “mechanism” covering how programs and apps causally interact, but that isn’t how they usually talk about their skills. Their knowledge resembles procedural routines more than programming skills. If you ask a teenager to give examples of how loops, queues, stacks, and arrays interact in simple programs, you may be met with a blank and sometimes hostile stare. Most teenagers’ coding skills don’t seem to be symbolic equivalents of advanced mechanical understanding. If they were, companies would not be madly scrambling to find even marginally competent programmers. Today’s youth, the so-called digital natives, often seem to be gliding along on the surface of a vast system that they do not understand at all. They may have remarkable “native” skills at using and learning to use software-intensive interfaces, but if the real programmers have done their jobs well, the actual workings of those devices may be not at all transparent to those mastering the user interface. Software- shuffling skills at the user level may be completely different from knowing how and why. Teens may know that graphics processing unit chips are better for machine learning but not how they speed up processing. These gaps in understanding will only get worse as deep learning systems solve more and more problems for us. If a company comes up with an improved facial recognition system, people are unlikely to ask how its mechanism has changed. Even experts in the area just assume it arises from a more powerful deep learning system, which is a black box resistant to any useful interpretable explanation of how it works.

In short, the world of artifacts has been radically transformed in the past fifty years. Readily apparent underlying mechanisms have vanished into mazes of circuits and blocks of silicon with millions, and often billions, of transistors. I remember when my father brought home a ten-transistor portable radio, describing it as a huge technology leap over our prior six-transistor… “
Above is the full text from page 99 of Wonder. It resonates surprisingly well with several of the core messages of the book: Well before children start formal schooling, they reveal a drive to uncover the mechanisms that underlie and give rise to all that they encounter. I describe this drive as a joyous urge to wonder “why” and “how”. Page 99 relates to this early emerging passion to know more about mechanisms. This passage links to another central theme – wonder becomes suppressed in most children around age five or six; and the decline of wonder may be worsening as fewer and fewer devices have revealing innards. Page 99 asks whether today’s children inhabit “mechanism deserts,” where insides no longer readily reveal how devices work. This is part of a broader discussion how wonder becomes stifled in children and may stay suppressed for a lifetime, resulting in disengagement from science and susceptibility to misinformation. But that future is not inevitable. Wonder illustrates how to avoid lives of impoverished understanding with barely any sense of the remarkable causal underpinnings of all that is around us. A large body of recent empirical research tells us what we can do as individuals and as groups to maintain and revive wonder at any age. Page 99 may not be the very best page for capturing the whole book, but it is certainly a good one.
Learn more about Wonder: Childhood and the Lifelong Love of Science at The MIT Press website.

--Marshal Zeringue