Cognitive Daily

How we learn to walk

Learning to walk was a passion for my son Jimmy. He would sweat and struggle with it until finally he had it mastered—and then it was off to the races. My daughter Nora, by contrast, didn’t seem to mind not being able to walk. After all, if you didn’t walk, then some sweet grown-up would soon show up and carry you wherever you wanted to go. The photo below illustrates another way Nora convinced others to do the walking for her:


Nora apparently was oblivious to the fact that parents, grandparents, aunts, and uncles around the world watch infants’ progress in walking with anxious anticipation—all the while dreading the thought that little Johnny might not learn to walk as quickly as little Joey across the street.

For nearly a century, psychologists have studied infant walking and developed a seemingly infinite variety of explanations about when infants learn to walk. While they’ve typically agreed that babies will not learn to walk until they have sufficient strength and balance, they’ve disagreed about what causes these two factors to develop. One hypothesis holds that it is the physical maturation of the body that enables walking: once legs are long enough and the chubby baby fat yields to muscle, then walking naturally follows. Another explanation suggests that the key is the development of the brain. The brain increases from 30% to 70% of its adult size by age 2. Perhaps it’s this phenomenal growth of brain power that allows kids to manage to balance themselves for those critical first few steps. A third possibility is that learning to walk simply takes practice, and the more practice a baby has, the sooner she will be able to walk proficiently.

Karen Adolph and Patrick Shrout of New York University and Beatrix Vereijken of the Norwegian University of Science and Technology designed a study to try to determine which of these theories is most promising (“What Changes in Infant Walking and Why,” Child Development, 2003). They developed a simple method to record the footsteps of hundreds of infants, as well as a few kindergartners and adults to compare them to. The method was decidedly low tech: they attached moleskin pads to the walkers’ shoes (a triangle at the toe and a square at the heel), then inked them up with different colors for each foot and had them walk the length of a sheet of butcher paper. The marks provided a permanent record of each walker’s path.

(As a demonstration of the difficulty of conducting experiments with infants, the researchers had to use a few tricks to get them to complete the task. The babies walked across the top of a long platform, raised about 3 and a half feet, so that there was no possibility of their running away. For safety, the experimenter walked alongside the infants, who were called by their parents at the end of the platform. What’s more, the data for 11 infants had to be scrapped because they refused to hold still to have their bodies measured.)

The researchers carefully tracked age and the start of walking for each of the infants, and measured the leg length, weight, and Ponderal Index (a measure of “chubbiness”) for all participants. The youngest babies took the shortest steps and had the largest step width. As they became better walkers, their steps became narrower and longer compared to their leg length. There were no significant differences in walking ability between kindergartners and adults.

Adolph et al. then cross correlated the data they had compiled with all the different developmental factors they had measured. None of the physical measures—Ponderal Index, leg length, weight, and so on, made a significant contribution to the variability in infant walking ability. Even the age of the infant was not a significant factor. The only significant contributor to walking ability was days of practice, which was determined simply by the number of days since the child was first able to walk.

The researchers reasoned that physical maturity or brain development could not be responsible for the ability to walk. The fact that walking begins at a wide variety of ages means that it is probably not related to the development of the brain: babies’ brains simply don’t mature much differently from one another. Once a baby decides to try to start walking, he or she progresses just about as rapidly as any other child, based on how much the baby practices.

This model certainly explains the difference between Jimmy and Nora’s learning to walk. Nora simply cared less about walking, and so chose to learn how to do it later than her brother. Once she decided to do it, she learned to walk as well as anyone. Jimmy was an early walker, but he didn’t learn how to do it any faster or slower than anyone else.


  1. #1 J.D. Fisher
    June 7, 2005

    Great study. I understand that the Ponderal index is birth weight x birth length x 100. Is this correct?


  2. #2 Dave Munger
    June 7, 2005

    The Ponderal Index is a measure that can be taken at any time, not just birth, and is equal to weight divided by height cubed (not sure how to format that in HTML—sorry!).

  3. #3 Tom Murphy
    June 7, 2005

    How is the Ponderal Index different from a person’s BMI?

  4. #4 Dave Munger
    June 7, 2005

    Well, the BMI is essentially weight divided by height squared, so that’s the main difference. Ponderal is weight divided by height cubed.

  5. #5 Mind Hacks
    June 10, 2005

    Spike activity 2005-06-10

    Quick links from the past week in mind and brain news: A widely reported story suggests that Ashkenazi Jews may be genetically more likely to be highly intelligent. Full text of research paper here. Psychometrics, the science of measuring the mind, has…

  6. #6 Patrick West
    November 4, 2005

    Im 15 and in my health class, we tried to get the Ponderal Index of randon weights but alot of the kids thought it was confusing, is there an easier way to get the Ponderal Index?

  7. #7 Dave Munger
    November 4, 2005

    Well, here’s a link to a Ponderal Index calculator. I can’t vouch for its accuracy, but it seems about right.