Interesting Findings, Need More Testing.

This 1982 study by Clark and Hecht answers a critical question concerning language acquisition, but in doing so it raised some questions of my own. That's science, I guess: the more you know, the more you know you don't know.

In any case, this study set out to affirm a specific sequence for language acquisition in humans concerning -er suffixes. Their findings are summarized in the conclusion section:

"We have argued in this paper that children fill such gaps by relying on certain principles that guide their acquisition of a repertoire of word-formation devices. Initially, the principle of semantic transparency guides the children by giving priority (a) to familiar words in constructing new word-forms and (b) to the use of one form or one meaning. A second principle, productivity, allows the child to identify more specialized word-formation devices to convey particular meanings. And a third, conventionality, gives priority to conventional forms over other forms currently in the child's repertoire." (22)

While I don't disagree with any of these findings, I am a little skeptical of them given the nature of the study that arrived at the conclusions. My concerns can be divided into 3 major areas:

• For one, the study featured a sample size of only 48 children, which seems a little small in comparison to other linguistic studies we have looked at this quarter. This seems like a sample size for a study that confirms conventional wisdom, rather than a sample size for a study that tries something new.
• All subjects went to the same school, where it is reasonable to think that language is taught the same way. Because of this, it is hard to account for the role education plays in language acquisition, and how different methods of teaching could result in different sets of principles.
• All subjects come from similar language backgrounds: English as a first language. Because of this, I don't believe it's appropriate to extend these conclusions to other languages, such as German; Clark and Hecht themselves acknowledge a study by Stern and Stern (1928) that suggests that German-speaking children may develop understanding about agent and instrument nouns before their English-speaking counterparts.

For these reasons, I would like to see a study in which

• More children are analyzed,
• The children come from separate schools,
• The children don't speak English as a first language.

If such a study yields similar conclusions as Clark's and Hecht's, it would do wonders to the study's legitimacy. Quite frankly, given that Clark and Hecht conducted this study over 30 years ago, I wouldn't be surprised if such a study already exists,