Past Projects

Previous research in the Little Learners Laboratory was guided by an overarching interest in the acquisition and organization of knowledge throughout early childhood. Much of that work focused specifically on interactions between categorization, word learning, and conceptual knowledge that, from my perspective, dynamically interact to support development. As such, this research sits at the heart of sustained theoretical controversy regarding children’s capacity for accessing conceptual knowledge in a variety of learning contexts. Some specific examples of projects completed in the LLL are described below.

 

Conceptual supports for learning

In our first foray into this area of research, Booth (2009) demonstrated that words are more easily learned by 3-year-olds when descriptions of the causal properties of their referents were provided than when descriptions of their non-causal properties were offered instead. With funding from the National Science Foundation (#0843252), we proceeded to ask a number of more specific questions about this effect. First, we attempted to better articulate the mechanisms involved. Booth (in press) showed that the influence of causal information is evident during the initial acquisition phase of learning, suggesting that it enhances attention at the time of encoding. Our results further suggest that causal information facilitates initial acquisition, at least in part, because it is inherently interesting. Specifically, preschoolers generally prefer to learn about causal over non-causal information (Alvarez & Booth, 2011), and children’s desire to learn about the causal properties of novel items is sufficiently strong to motivate their sustained engagement in an otherwise boring task (Alvarez and Booth 2014). In addition to its apparent influences on attention and motivation, causal information might also contribute to better learning by providing a framework for the coherent elaboration of semantic representations. Consistent with this possibility, richly-causal categories better support inductive inference in preschoolers than do weakly-causal categories (Booth, 2014).

In other related lines of research, we are mapping the developmental trajectory of the influence of causal information on early word learning and are exploring the relevance of causal information to facilitating learning in more naturalistic learning contexts. Thus far, it appears that the influence of causal information on word learning is maintained through kindergarten in traditional laboratory settings, but begins to decline in first grade (Booth, 2014) and in otherwise more stimulating settings (e.g., picture-book reading and group project play). We suspect that these results might be related to variation in the motivational structures engaged at different ages and in different contexts. For example, we suspect that reductions in the influence of causal information on word learning as children age may be related to well documented declines in intrinsic motivation that accompany the accumulation of formal schooling.

Object function

Object function is perhaps the instantiation of conceptual information most widely accepted in the literature. Early work in the LLL (Booth, 2001; Booth & Waxman, 2002) demonstrated that object function facilitates categorization in infants as young as 14 months of age. This finding suggests not only that this form of conceptual information is accessible and salient to infants, but also that it is instrumental in supporting a fundamental cognitive process (i.e., categorization). In subsequent work funded by NIH (RO3 HD048759-01) and NSF (0445871) we addressed several more specific questions regarding the role of function in supporting early categorization. For example, Booth (2006) demonstrated that object function exerts its effect by both enhancing infants’ attention indiscriminately to all commonalities among objects, and by training infants’ attention more specifically on just those object features that are causally related to object functionality. Booth, Schuler and Zajicek (2010) further demonstrated that function enhances the efficiency with which infants form new categories by reducing the need for comparison across exemplars. That is, infants were more likely to detect categories after exposure to only a single exemplar if they had also observed the function of that exemplar, than if they did not. Booth (2008) then began to specify what it is about object function that supports its facilitative effect on categorization, and identified the causal relations connecting objects to interesting outcomes as an important factor.

In later work, Elizabeth Ware (Ware & Booth, 2010 & Booth & Ware, 2010) demonstrated a role for object function in supporting the emergence of the ‘shape bias’ as a constraint on early word learning. Six weeks of experience with categories embodying tight causal relationships between object shape and function led to a precocious shape bias in toddlers. Non-functional experience with the same categories did not. This finding suggests that the shape bias stems from children’s appreciation of conceptual links between shape and less perceptually obvious properties, like function.

 

Knowledge of ontological kind

In other work, we focused on children’s ontological knowledge as another form of conceptual information that might affect early cognition. In Booth and Waxman (2002), we reported that 3-year-olds extend novel words on the basis of similarities in shape alone (i.e., they demonstrate a ‘shape bias’) when labeled objects are described as having conceptual properties typically associated with artifacts (e.g., has a special use) while children extend labels on the basis of similarities in both shape and texture when the very same objects are described as having conceptual properties typically associated with animate kinds (e.g., has emotions). Because the objects presented in both conditions were perceptually identical, only the domain-specific conceptual information provided could be responsible for the observed differences in extension patterns. I have since extended this finding to 2-year-olds, and have even shown similar effects in 18- to 22-month-old infants (Booth, Waxman & Huang, 2005, Booth & Waxman, 2003, 2006 and 2008).

 

Socio-pragmatic knowledge

In other work, we attempted to tease apart the influences of attentional capture and conceptual knowledge of communicative intent on toddler’s use of gestural cues to disambiguate novel word reference (Booth, McGregor and Rohlfing, 2008). We began by recording the looking patterns of toddlers while they heard words applied to novel objects by a speaker who looked at, look at and pointed to, looked at, pointed to and touched, or looked at, pointed to, touched and manipulated the intended referent. Learning (as assessed through subsequent forced-choice comprehension testing) improved significantly with greater redundancy among cues, with the largest improvement evident when pointing was added to gazing. Looking times revealed that attentional factors accounted for only a small fraction of the variance in performance, thereby suggesting a strong role for socio-pragmatic factors (e.g., conceptual knowledge of the link between communicative intent and pointing) in guiding performance on the task.

We also completed a similar project with children diagnosed with autism (Patrick, Hurewitz & Booth, 2013). This population is known to have considerable difficulty understanding the mental lives of others and the impact of mental states on behavior (i.e., theory of mind). As a result, one might expect the word-learning difficulties observed among children with autism to be, at least in part, explained by their inattention and/or inability to interpret the gestural cues of communicative partners. Our findings are consistent with this possibility. Young children with autism spent significantly less time attending to the speaker during labeling episodes, and the amount of time that they did so predicted their word-learning performance.