Adapting to children: Information redundancy in language production

During conversations, speakers make decisions both about what they want to convey and how to communicate that with language. Previous research established that speakers are biased toward efficient communication: being sufficiently informative for listeners, while minimizing their own production effort (Grice, 1975; Jaeger & Buz, 2018; Zipf, 1949). Hence, efficient speakers expend less effort producing redundant elements (i.e., those typical of the situation and predictable with real-life knowledge). For example, Westerbeek et al. (2015) found more mentions of the colour blue over red when referring to tomatoes, thereby reducing redundancy since tomatoes are stereotypically red. This creates an interesting situation when speaking to children. Unlike adult interlocutors, who have already acquired sufficient real-world knowledge, children are still learning what is typical in the world. Therefore, what is redundant for an adult may not be such for a child. Consequently, when talking to children, the pressure for efficiency may be outweighed by the need to use language informatively to explain how the world is, thereby leading speakers to produce signals that would be considered to be redundant for adults in order to maximize the informativity for a child. Tal et al. (2021a, 2021b) showed that speakers used longer references like full noun phrases for highly predictable targets when talking to younger compared to older children, and this redundancy declined as children aged.