Complex sentences are just that; complex! Although that may seem self-evident, nothing hammers home the intricacies and complexities in everyday language quite like studying complex sentences in naturalistic data. This was the focus of the past six weeks of my summer after I was awarded a LuCiD internship grant.
During my undergraduate studies, I have been lucky enough to have been involved in child language research at Cardiff University and here in Manchester. Whilst this past experience in child language focused upon the role of context, this LuCiD summer studentship allowed me to work on a project studying complex child language from a more grammatical perspective. Given my aspirations toward a career in research, the chance to gain insight into another side of child language has been invaluable, highlighting just how many facets of research exist within a single area.
The focus of the project was the occurrence of complex sentences in naturalistic data, in particular the subordinators before, after, because and if. A large body of research indicates that temporal subordinators such as before and after present children particular difficulty (Amidon & Carey, 1972; Blything, Davies & Cain, 2015; Clark, 1971; Johnson, 1975). As this difficulty is not found in other ordered tasks (such as “do this first, do that last”) (Amideon & Carey, 1972), it seems that it is the relational aspect that children experience difficulty with. Interesting questions are also posed around children’s burgeoning understanding of causality and how this is represented in their speech, with suggestions that causal structures are originally used to justify their own speech acts (Kyrantzis, Guo & Ervin-Tripp, 1990).
How children begin to understand these terms, and decide which command to follow, has long been a matter for debate. It has been suggested that children focus on the main clause, rather than the subordinate, when interpreting temporal sentences (Amidon & Carey, 1972). Others claim that comprehension of such sentences is reliant upon whether the order that events are mentioned in the sentence (order of mention) is the same as order of occurrence (Clark, 1971). Regardless of complex subordinator in question, the question remains: are children naturally better at dealing with whatever sentence structure they hear the most often?
To investigate this, my internship involved looking at naturalistic data in order to ascertain the frequency of various aspects of complex sentences in children’s language. These would then be compared to the parent data in order to see whether children’s language follows the same patterns.
After performing a reliability check upon the parent data, I coded and analysed utterances from the child data. This allowed me hands-on experience with both corpora, highlighting the differences and similarities between the two. Most strikingly was the hyper-complexity in the parent data; with instances of 5 or 6 subordinated clauses within a single sentence, I was left astounded that we ever learn to speak at all! Comparatively, the child data was far simpler, however it presented its own challenges; with many utterances incomplete or unclear, omitting vital aspects such as subject and verbs.
Alongside analyzing this data, I was fortunate enough to go into schools and nurseries to view the experimental complement. This side of the research involved investigating children’s understanding of these clauses in tangent with other linguistic measures, as well as understanding of causality. How this understanding maps onto exposure is yet to be seen, however it undoubtedly holds important theoretical and applied implications for child language development. Understanding how children’s comprehension and production relate to one another gives valuable insight into the processes involved, informing our understanding of typical development.
More generally, gaining further insight into the role of language environment upon language production informs our understanding of the importance of exposure, with broad implications for theoretical standpoints of language development. Although many aspects of the way in which children begin to understand and use such complex structures is uncertain, this research will begin to uncover just when and what they hear – and whether this exposure predicts their own language use.