Imitate to Accumulate: The Relationship Between Syntactic Priming and Long-term Learning, 2019-2022

Children’s language closely reflects their recent and long-term experiences of language. Within conversations, children often repeat the words and sentence structures that they have just heard; their vocabulary and grammatical development tends to reflect the diversity and complexity of their caregivers’ language. But little is known about how children’s short-term language experiences contribute to their longer-term language learning. Syntactic priming effects may offer a promising explanation: growing evidence suggests such effects persist and accumulate to affect language use within the same interaction and even a week later. Accounts of syntactic priming as learning predict age-related differences in the magnitude of immediate priming and cumulative learning over multiple immediate experiences of syntactic structures which should lead to long-term changes in speakers’ representations of syntactic structures. This study investigates whether children’s behaviour at different stages of development supports these predictions.  We ran two experiments examining the timecourse of experience-based effects in children at early and later stages of acquisition and a comparison adult group. Both experiments involve two testing sessions, consisting of a relatively large number of items (N=48), separated by one week. Experiment 1 assesses priming of noun phrase (NP) structures where participants take turns in describing target pictures with an experimenter who alternates between adjective-noun (AN: a blue cat) and noun-relative clause (RC: a cat that’s blue) primes. Experiment 2 tests verb phrase (VP) structures (specifically actives (a cat chased the dog) vs passives (the dog was chased by a cat)). We predicted that all groups will show immediate priming effects within sessions such that participants will produce more target structures after the same prime than after the alternative prime. We also predicted long-term effects of experience, such that participants will be more likely to produce target structures in Session 2 than Session 1. Moreover, we expected younger children to show larger immediate priming effects than older children or adults, leading to greater long-term learning effects for children at earlier stages of acquisition. Both experiments show patterns that are consistent with immediate priming effects at all ages, but though children are showing the largest effects, it is not necessarily only in the youngest age group. Contrary to expectations, the data suggest that there could be either no difference, or a decrease in priming across sessions.

How we learn and use language is, not surprisingly, related to the language we experience around us: ultimately children who are exposed to English learn English, but more specifically, research shows that children exposed to varied language input (wider vocabularies, diverse sentences) come to develop more extensive language skills than those with narrower input. While we know that children's experience with language is important for shaping their learning of language, it remains unclear precisely how our experience with language influences our language development: what aspects of language experience are important, and how do children make use of them? Our project investigates how children learn from their language experiences, and the underlying learning mechanisms that they use to do so. Understanding the mechanisms that support language learning is theoretically important, because it informs our understanding of a uniquely human ability. But it also has substantial societal implications, because successful language development is critical to later educational and social attainment. Our project focuses on how language experience affects children's use of particular grammatical or syntactic structures - the way words are ordered in a sentence. We know that one way in which both children and adults make use of the language they experience is by re-using it. This re-use occurs not just for specific words (e.g., saying 'sofa' instead of 'settee' because that's what your friend just called it), but also at more abstract levels, including grammar. For instance, you are more likely to say "those pictures were drawn by Quentin Blake" after hearing someone say "this book was written by Roald Dahl". This grammatical repetition is known as syntactic priming: hearing a word order makes it easier for you to re-use that order, even with different words. When a speaker repeats a structure they have been exposed to, it indicates that they have a mental representation for that structure that they can use when understanding a sentence, and then re-use when planning a new sentence to say. So patterns of syntactic priming effects in children provide evidence about which grammatical structures they know. More recent research suggests that the influence of such priming lasts longer than the immediate context. In fact, recent models of language processing propose that priming can lead to long-term learning, so that when someone experiences a grammatical structure, it sets in train a persistent change in their language representations. For example, when a child hears "this book was written by Roald Dahl", it does not just help them to re-use that structure immediately, it also strengthens their knowledge of that structure. But it is unclear how this learning occurs: how does children's immediate experience with language translate into longer-term development of adult-like representations? In this research we will conduct psycholinguistic experiments with children at different stages of language development in which we vary which types of structure children hear and say, in different contexts and at different time points. We will examine how likely they are to re-use these structures and how long-lasting such effects are, in order to cast light on the way in which children can learn from different elements of the language they experience. We will address the following questions: When in language development is this type of experience important? To what extent is experience-based learning linked to specific contexts, whether linguistic (experience of particular words) or non-linguistic (experience within a particular conversation or task), and to what extent does it generalise? Does learning occur to the same extent through the act of hearing different word orders and the act of saying them? Answering these questions will ultimately help us understand the mechanisms, conditions and trajectories underlying successful language development.