Children produce first word combinations and sentences around 18 and 21 months, respectively. By 24 months, grammatical forms begin to appear that mark sentences for tense and agreement (e.g., "There's a car." "This goes here."). Throughout the third year, children vary in their production of sentences. Sometimes they produce sentences without a tense morpheme like "Daddy _ working", and at other times, they produce the adult-like version "Daddy is working." With development, children fully integrate these tense and agreement morphemes into a generative mental system. In this sense, generativity refers to children's ability to produce new sentences never heard before with the expected tense and agreement morphemes. This project measures developmental change in this generative system and tests the hypothesis that grammatical growth is characterized by a predictable developmental trajectory with individual variation surrounding this trajectory. A productivity approach is used to measure grammatical development to minimize the likelihood that early sentences are produced as part of automatic phrases. Automatic phrases can be produced without using underlying grammatical knowledge in much the same way that a second language user learns to produce a few stock phrases like "si se puede." Inclusion of automatic phrases or "direct activations" has obscured accuracy measures of tense/agreement morpheme use as well as efforts to characterize patterns of developmental change under the age of 3. This project has three specific objectives: (a) to reveal the measurement error introduced by direct activations at the onset of grammatical development, (b) to characterize general patterns and individual variation in the growth of tense and agreement; and (c) to test the stability of grammatical growth during this early period. Language samples from 50 children obtained at 21, 24, 27, 30, 33 and 36 months will provide the primary data for modeling the growth of tense and agreement. Study 1 will compare growth models using productivity versus accuracy measures. Study 2 will use data from these models to predict 36-month tense/agreement accuracy on standardized probes. Study 3 will compare the extent to which early productivity and accuracy growth trajectories (21 to 27 months) predict later accuracy growth trajectories (27 to 36 months). Finally, Study 4 will determine the extent to which the frequency, informativeness, and diversity of tense/agreement morpheme use in parental language input at 21 months explains the variation in growth rates. The results are expected to reveal both the gradual nature of grammatical growth and the existence of robust individual variation. In addition, the findings will address the contribution of language input to growth and variation.
Theories of grammatical development must explain the variable use and gradual acquisition of tense and agreement. Current theoretical debate has focused primarily upon why variable use occurs. The current project will refocus the debate upon why some children resolve this period of variable use more rapidly than others. Differences in developmental rate require explanation and are likely to be affected by a complex combination of biological, environmental, and developmental factors. However, improved measurement precision and documentation of underlying growth trajectories are critically needed as a foundation for these future investigations. In addition, documentation of the typical range of variation for children under the age of 3 will provide developmental expectations that can be used to improve the early identification of children at risk for persistent language impairments.
from its very onset. The tense and agreement system is a microcosmic example of grammatical development in general. We have learned that the tense agreement system is not present at 21 months of age and that it grows in strength and productivity with a gradually accelerating rate over the course of the third year of life. But there is also a large amount of individual variation between children in the rate at which the system grows and this variation has allowed us a new window into the mechanisms of the acquisition of grammar. The project has shed light on two components of early grammatical development: (a) cross morpheme facilitation; and (b) naturally occurring transitional probabilities in language input. The former is a learning process that occurs within the child, aiding the generalization of grammatical knowledge. The latter is a characteristic of input to children that can promote or hinder the rate of grammatical growth. Broader Impacts: Quantification of individual variation in grammatical growth rates will form the foundation for future investigations of the biological, environmental, and developmental predictors of these differences in rate. Documentation of general growth patterns and the typical range of variation provide developmental expectations for children under age three that improve early identification of young children at-risk for language impairments. The investigation of language input effects undertaken in this project will help provide an empirical foundation for future language intervention with those children.
