Roman Feiman studies how children learn to reason, express their reasoning in words, and understand the reasoning of other people. To study how young children first learn to reason, he focuses on simple logical concepts and words, like “no” and “or”. He studies what makes learning such words hard, and what nevertheless allows children to do it successfully.
My plans for the fellowship period
During my fellowship, I plan to investigate how the structures of different languages can make it easier or harder for children to learn how to express their reasoning. Some languages, like English, have two abstract words to express negation – “no” and “not”. Others, like Spanish, have only one word for negation. Still others, like Hungarian, Hebrew, and Tagalog, use a variety of words to express refusing, prohibiting, denying, or saying that something has disappeared. Does having more different words make learning each meaning easier – or harder? Do children learn how to express their reasoning with these words differently in different languages? Looking at what children who are learning different languages say will help tell how their ability to express different meanings changes over development, and how this might depend on their particular language. Looking at how children understand different meanings will help tell how they develop verbal reasoning skills, which they will rely on in their formal education — in any language.
How will my work change children’s and youth’s lives?
Schools ask children to solve verbal reasoning problems in every subject, at every level of education. This means that to learn specific content knowledge, children depend on a general ability to understand logical words and concepts; If we drop a rock and a feather, will the rock or the feather land first? If the rock does not land first, why?
I aim to understand the influence of language structure on the development of verbal reasoning and on variability in its use in further learning. By separating linguistic from cognitive challenges, my work promises to separate variance between languages from variance in cognitive factors, which are dependent on SES background and learning environments. Differences in achievement that may appear to be due to differences in educational or economic factors might turn out to be due to differences between the structures of languages.