When your child hears a new word, do they first check whether the speaker is reliable? And what do they do when a new word clashes with a name they already know? Researchers Shalini Banerjee and Ishanti Gangopadhyay of Indiana University asked exactly this in a study published in PLOS ONE in 2026.
They studied 4- to 7-year-old children who spoke only English. The article text available to us does not state how many children took part. Each child saw pairs of objects, one unfamiliar and one familiar. A speaker then taught them new names. Some speakers had earlier named familiar objects correctly (the "accurate" speaker). Others had named them wrongly (the "inaccurate" speaker). The researchers then tested how well children remembered the new words over the short term.
New names for new objects stuck, whoever said them
In one version, the speaker gave a new name to the unfamiliar object. This matches a habit called mutual exclusivity: children tend to assume one object has one name, so a new word probably belongs to the thing they cannot yet name.
In this version, children remembered the new labels from both the accurate and the inaccurate speaker. The authors suggest children are open to learning words from speakers regardless of their past accuracy. They may also use other cues, such as which word and object were shown together.
The catch: a new name for something they already knew
In the other version, the speaker gave a new name to the familiar object, which clashed with the "one object, one name" expectation. Here, children performed poorly.
The authors say this is possibly because children found it hard to sort out what the speaker said from what they already knew. Their overall conclusion is that short-term word memory is supported when the available cues agree, and becomes less robust when they conflict.
What this study cannot tell us
- It measured short-term memory only, so we do not know whether the words lasted.
- The children spoke only English. The study did not cover bilingual children, and the text we have does not mention Singapore.
- The explanation for the poor results in the clash condition is a possibility the authors raise, not something they proved.
Small ways to line up the clues when teaching new words
This was a controlled experiment, not a home study, so treat these as gentle ideas rather than rules.
- Name an unfamiliar object when teaching a new word. Children remembered new labels best when the new name went with a new object.
- Do not worry that an earlier mix-up will block learning. Children learned from both accurate and inaccurate speakers when the cues agreed.
- Avoid giving a new name to something your child already knows by another name. Those labels were remembered less well.