Dare A. Baldwin’s research while affiliated with University of Oregon and other places

Once one sees a pattern, it is challenging to “unsee” it; discovering structure alters processing. Precisely what changes as this happens is unclear, however. We probed this question by tracking changes in attention as viewers discovered statistical patterns within unfolding event sequences. We measured viewers’ “dwell times” (e.g., Hard, Recchia, & Tversky, 2011) as they advanced at their own pace through a series of still-frame images depicting a sequence of event segments (“actions”) that were discoverable only via sensitivity to statistical regularities among the component motion elements. “Knowledgeable” adults, who had had the opportunity to learn these statistical regularities prior to the slideshow viewing, displayed dwell-time patterns indicative of sensitivity to the statistically defined higher-level segmental structure; “naïve” adults, who lacked the opportunity for prior viewing, did not. These findings clarify that attention reorganizes in conjunction with statistically guided discovery of segmental structure within continuous human activity sequences. As patterns emerge in the mind, attention redistributes selectively to target boundary regions, perhaps because they represent highly informative junctures of “predictable unpredictability.”

We identify three “working edges” for fruitful elaboration of the Chunk-and-Pass proposal: (a) accounting for the earliest phases of language acquisition, (b) explaining diversity in the stability and plasticity of different representational types, and (c) propelling investigation of action processing.

In the real world, causal variables do not come pre-identified or occur in isolation, but instead are embedded within a continuous temporal stream of events. A challenge faced by both human learners and machine learning algorithms is identifying subsequences that correspond to the appropriate variables for causal inference. A specific instance of this problem is action segmentation: dividing a sequence of observed behavior into meaningful actions, and determining which of those actions lead to effects in the world. Here we present a Bayesian analysis of how statistical and causal cues to segmentation should optimally be combined, as well as four experiments investigating human action segmentation and causal inference. We find that both people and our model are sensitive to statistical regularities and causal structure in continuous action, and are able to combine these sources of information in order to correctly infer both causal relationships and segmentation boundaries.

Past research has established that children typically learn better from live demonstrations than from two-dimensional (2D) media. In the present set of experiments, we investigated the efficacy of a new 2D learning medium—the self-paced slideshow. A primary goal was to determine whether the “video deficit effect” extended to self-paced slideshows. In Experiment 1, preschool-age children saw demonstrations of novel events either live, on video, or by advancing through self-paced slideshows. They were then tested on their performance and verbal memory. In line with past work, children in the live condition outperformed those in the video and slideshow conditions at reproducing the target actions. To further explore the 2D media, Experiment 2 directly compared learning from self-paced slideshows to that from videos. Changes to the stimuli included a more natural extraction rate of slides and a higher focus on the objects. Children’s performance differed little between conditions, with the exception of reproducing fewer actions in the slideshow than video condition on two (of four) toys. Ultimately, we conclude that the video deficit extends to self-paced slideshows. Future work must investigate how to enhance children’s learning from 2D sources, given their increasing role in daily life.

We investigated how processing unfolds across time as adults view novel manual activity. Adults advanced through slideshows of an actor performing sleight-of-hand manoeuvres; dwell times were recorded to each slide. As in past work involving larger-scale intentional action scenarios, adults allocated more attention to slides depicting breakpoints-transitions from one event segment to the next-than slides displaying mid-stream action, indicating that dwell times are sensitive to adults' recovery of segmental structure in intricate manual activity. Adults also allocated increased attention towards slides indicative of causal structure, indicating that dwell times index processing of causally relevant information as well as segmental structure. We were additionally curious if adults' learning outcomes would differ as a function of whether they paced themselves or watched slideshows advance at a computer-controlled pace. Memory did not differ, but adults viewing computer-controlled slideshows displayed an advantage in performing the tricks. A single attentional focus may have helped adults learn to re-enact the actions, relative to a split attentional focus.

We present a series of studies using a new methodology—the dwell-time paradigm—to investigate observers’ sensitivity to structure within unfolding action sequences. Viewers advance at their own pace through slides extracted from digitized videos; dwell times for each slide are recorded. Findings reveal that adults, preschoolers, and even infants display sensitivity to both segmental and hierarchical structure within events. Other work hints that dwell times may also reflect infants’ and adults’ sensitivity to violations of causal structure. In addition, social context (e.g., pedagogy) appears to influence both infants’ dwell-time patterns and their later successful enactment of a sequence. Although questions remain about the precise way to characterize dwell-time effects, the paradigm shows promise for providing altogether new information about action processing and how it changes with development.

Generic noun phrases, or generics, refer to abstract kind categories (Dogs bark) rather than particular individuals (Those dogs bark). How do children distinguish these distinct kinds of reference? We examined the role of one socio-pragmatic cue, namely pointing, in producing and comprehending generic versus particular reference. Study 1 demonstrated that parents of preschool-aged children pointed more when referring to particular instances versus generic kinds. Studies 2 and 3 addressed how children interpreted pointing when linguistic cues were ambiguous with respect to the generic versus particular distinction, for example, They are afraid of raccoons said in the presence of several dogs, where they could refer to the generic category (dogs) or a particular set (the/those dogs). Results indicate only a partial socio-pragmatic sensitivity to pointing's role in marking particular reference. They additionally speak to issues related to children's acquisition of generics and their expectations regarding transmission of generic knowledge.

Recent developmental work demonstrates a range of effects of pedagogical cues on childhood learning. The present work investigates natural pedagogy in informal parent-child play. Preschool-aged children participated in free play and a toy task with a parent in addition to a toy task with an experimenter. Sessions were extensively coded for use of pedagogical cues, such as eye contact and pointing. We present a series of analyses investigating the pedagogical cues that characterized natural pedagogy, how these cues related, and how cues bundled into facets. Implications for future research and determining the validity of these measures of natural pedagogy are discussed.

We investigated infants' response to pedagogy in the domain of tool use. In experiment 1, infants viewed a causally relevant tool-use demonstration presented identically in either a social/pedagogical or social/non-pedagogical context. Infants exposed to pedagogical cues displayed superior production of the tool-use sequence. This was so despite infants displaying equivalent attention to the demonstration across conditions. In contrast, pedagogical cues had no systematic impact on infants' discrimination between causally possible vs. impossible tool-use sequences in a looking-time task. Interestingly, however, older infants across both conditions displayed a preference for looking toward the causally possible display. Experiment 2 documented that social cues of any sort (regardless of pedagogy) accompanying the demonstration triggered older infants to discriminate the causally possible vs. impossible events whereas a non-social demonstration did not. Together, the two experiments implicate ‘social gating’ as well as a pedagogical stance in infants' processing and execution of causal action.

The current study addressed the degree to which maternal speech and action are synchronous in interactions with infants. English-speaking mothers demonstrated the function of two toys, stacking rings and nesting cups to younger infants (6-9.5 months) and older infants (9.5-13 months). Action and speech units were identified, and speech units were coded as being ongoing action descriptions or nonaction descriptions (examples of nonaction descriptions include attention-getting utterances such as “Look!” or statements of action completion such as “Yay, we did it!”). Descriptions of ongoing actions were found to be more synchronous with the actions themselves in comparison to other types of utterances, suggesting that: 1) mothers align speech and action to provide synchronous “acoustic packaging” during action demonstrations; and 2) mothers selectively pair utterances directly related to actions with the action units themselves rather than simply aligning speech in general with actions. Our results complement past studies of acoustic packaging in two ways. First, we provide a quantitative temporal measure of the degree to which speech and action onsets and offsets are aligned. Second, we offer a semantically based analysis of the phenomenon, which we argue may be meaningful to infants known to process global semantic messages in infant-directed speech. In support of this possibility, we determined that adults were capable of classifying low-pass filtered action- and nonaction-describing utterances at rates above chance.