What Toddlers Know They Don’t Know About Plurals

In 1958, psychologist Jean Berko Gleason developed her ingenious Wug Test to determine whether children between the ages of four and seven could extend the rules of English morphology (that is, how units of meaning such as stems and affixes can be combined) to new words. In the now-famous example, children were shown a sketch of a single, flame-shaped bird. “This is a wug,” they were told. “Now there is another one. There are two of them. There are two _____.” The children, much as you or I would,  responded, “wugs.” Explained Gleason in an interview decades later, “the Wug study showed clearly that young children do have rules: they know more than memorized individual words and this is shown by the fact that they can inflect words that they have never before heard, and thus could not have memorized.”

In the years since, psycholinguists have learned that children first begin producing plural forms for very common nouns quite early—by around a year and a half. But this doesn’t mean that children understand the plural morpheme –s, or any of its irregular pals, at an abstract level. Indeed, children don’t begin to comprehend plural markers on new words, like wugs, until age three. Work by my colleagues Cindy Lukyanenko and Cindy Fisher suggests that by this age children can also use principles of subject-verb agreement (“Where is…” vs. “Where are…”) to predict whether an upcoming noun will be singular or plural.

But what about before the age of three, during that 18-month gulf that separates earliest productions from relatively mature use (a period of time that is, quite frankly, full of bewildering numerical developments)? New experiments find that two-year-olds are aware of a singular-plural distinction in their language: we just have to look at what they don’t do to see it.

In research published in the Journal of Child Language in 2009, Jennifer Zapf and Linda Smith, psychologists at the University of Wisconsin-Green Bay and Indiana University respectively, presented 26 two-year-olds with a variation on Gleason’s Wug Test. First the toddlers were shown a picture of one new object (an obscure household tool, for instance), and provided a label (“This is a wug. Can you say ‘wug’?”). Then, they were shown a second picture. For half of the children this picture contained two wugs, and for the other half it contained a wug paired with a different, unnamed, novel object. “Can you tell me what’s on this page?” the experimenter asked. “What do you see?”

When presented with two wugs, the two-year-olds only produced the plural form “wugs” about 10 percent of the time—not surprising, given that nobody really expected them to succeed at this age. But strikingly, they didn’t produce the singular “wug” very often either. No, the toddlers’ most frequent response was to provide no response at all: on nearly 60 percent of trials in which a plural noun was appropriate, children refused to give an answer, compared to just a third of the trials in which a singular form was appropriate. A follow-up study revealed that two-year-old native speakers of Japanese—a language without mandatory plural marking—showed no such reluctance to extend the label for “one” to “more than one.”

In other words, the English-speaking but not the Japanese toddlers knew that they could not simply parrot back what they’d heard earlier. They knew that in their language the number of objects involved usually changed how the word was said. All they didn’t know, it seems, was how.