Gedeon O. Deák

Combining embodied models and empirical research for understanding the development of shared attention

The capacity for shared attention is a cornerstone of human social intelligence. We propose that the development of shared attention depends on a proper interaction of motivational biases and contingency learning mechanisms operating in an appropriately structured environment. Atypical contingency learning leads to deficits in shared attention as seen in children with autism. To test this theory, we describe a research effort that combines theoretically rigorous modeling techniques using both simulated and robotic learning systems with empirical investigations of social learning and development in infants and toddlers with and without developmental disabilities. We believe that studying embodied learning models, whose input data (from a real or virtual caregiver) is modeled after real infant-caregiver interactions, will lead to a better understanding of the development and dysfunction of shared attention.

Do children really confuse appearance and reality

Our understanding of many mental, social and physical phenomena hinges on a general understanding that appearances can differ from reality. Yet young children sometimes seem unable to understand appearance-reality dissociations. In a standard test, children are shown a deceptive object and asked what it really is and what it looks like. Many preschool children give the same answer to both questions. This error has been attributed to childrens inflexible conceptual representations or inflexibility in representing their own changing beliefs. However, evidence fails to support either hypothesis: new tests show that young children generally understand appearance-reality discrepancies as well as fantasy-reality distinctions. These tests instead implicate childrens failure to understand the unfamiliar discourse format of the standard test. This misunderstanding might reveal a subtler difficulty in making logical inferences about questions.

The law of large numbers in children's diversity-based reasoning

Adults increase the certainty of their inductive inferences by observing more diverse instances. However, most young children fail to do so. The present study tested the hypothesis that childrens sensitivity to instance diversity is determined by three variables: ability to discriminate among instances (Discrimination); an intuition that large numbers of instances increase the strength of conclusion (Monotonicity); ability to detect subcategories and evaluate numerical differences between the subcategories, or Extraction. A total of 219 Chinese children aged 6 to 11 were tested for sensitivity to diversity by means of Discrimination, Monotonicity, and Extraction. The results indicated that children at all ages were able to discriminate instances and attend to set size. However, only 9- and 11-year-olds demonstrated Extraction and sensitivity to diversity. Furthermore, among all children diversity scores increased linearly with the level of Extraction. These results suggest that the law of large numbers plays a role in childrens diversity-based reasoning.

A robotic model of the development of gaze following

For humanoid robots, the skill of gaze following is a foundational component in social interaction and imitation learning.We present a robotic system capable of learning the gaze following behavior in a real-world environment. First, the system learns to detect salient objects and to distinguish a caregiverpsilas head poses in a semi-autonomous manner. Then we present multiple scenes containing different combinations of objects and head poses to the robot head. The system learns to associate the detected head pose with correct spatial location of where potentially ldquorewardingrdquo objects would be using a biologically plausible reinforcement learning mechanism.

Visual Prediction in Infancy: What is the Association with Later Vocabulary?

Young infants can learn statistical regularities and patterns in sequences of events. Studies have demonstrated a relationship between early sequence learning skills and later development of cognitive and language skills. We investigated the relation between infants’ visual response speed to novel event sequences, and their later receptive and productive vocabulary. Using a modified visual expectancy paradigm (VExP; Haith, Wentworth, & Canfield, 1993), we tested 6-month-old infants’ speed at responding to novel but predictable contingent event sequences. In addition, parental reports and behavioral measures of infants’ vocabulary were obtained at 12, 16, and 22 months. In order to estimate the separate effects of linguistic input on vocabulary, maternal speech from a play session at 12 months was analyzed for lexical diversity and quantity. Results suggest that infants’ speed of responding to novel but predictable events at 6 months robustly predicted both receptive and productive vocabulary at 22 months. This relation cannot be attributed to general cognitive maturity, as measured by a standardized test (Bayley Scales of Infant Development; Bayley, 2005). Maternal input predicted additional unique variance in infant processing speed. The results suggest that infants’ capacity to quickly learn and respond to sequential patterns, over and above the quality of the speech input they receive, contributes to vocabulary size in the second year.

Editorial: New trends in Cognitive Science: Integrative approaches to learning and development

A new trend in Cognitive Science is the use of artificial agents and systems to investigate learning and development of complex organisms in natural environments. This work, in contrast with traditional AI work, takes into account principles of neural development, problems of embodiment, and complexities of the environment. Current and future promises and challenges for this approach are defined and outlined.

Methodological Considerations For Investigating the Microdynamics of Social Interaction Development

Infants are biologically prepared to learn complex behaviors by interacting in dynamic, responsive social environments. Although the importance of interactive social experiences has long been recognized, current methods for studying complex multimodal interactions are lagging. This paper outlines a systems approach for characterizing fine-grained temporal dynamics of developing social interaction. We provide best practices for capturing, coding, and analyzing interaction activity on multiple -temporal scales, from fractions of seconds (e.g., gaze shifts), to minutes (e.g., coordinated play episodes), to weeks or months (e.g., developmental change).

A Unified Account of Gaze Following

Gaze following, the ability to redirect ones visual attention to look at what another person is seeing, is foundational for imitation, word learning, and theory-of-mind. Previous theories have suggested that the development of gaze following in human infants is the product of a basic gaze following mechanism, plus the gradual incorporation of several distinct new mechanisms that improve the skill, such as spatial inference, and the ability to use eye direction information as well as head direction. In this paper, we offer an alternative explanation based on a single learning mechanism. From a starting state with no knowledge of the implications of another organisms gaze direction, our model learns to follow gaze by being placed in a simulated environment where an adult caregiver looks around at objects. Our infant model matches the development of gaze following in human infants as measured in key experiments that we replicate and analyze in detail.

Young children’s fast mapping and generalization of words, facts, and pictograms

To test general and specific processes of symbol learning, 4- and 5-year-old children learned three kinds of abstract associates for novel objects: words, facts, and pictograms. To test fast mapping (i.e., one-trial learning) and subsequent learning, comprehension was tested after each of four exposures. Production was also tested, as was childrens tendency to generalize learned items to new objects in the same taxon. To test for a bias toward mutually exclusive associations, children learned either one-to-one or many-to-many mappings. In Experiment 1, children learned words, facts (with or without incidental novel words), or pictograms. In Experiment 2, children learned words or pictograms. In both of these experiments, children learned words slower than facts and pictograms. Pictograms and facts were generalized more systematically than words, but only in Experiment 1. Children learned one-to-one mappings faster only in Experiment 2, when cognitive load was increased. In Experiment 3, 3- and 4-year-olds were taught facts (with novel words), words, and pictograms. Children learned facts faster than words; however, they remembered all items equally well a week later. The results suggest that word learning follows non-specialized memory and associative learning processes.

Category label effects on Chinese children's inductive inferences: Modulation by perceptual detail and category specificity

Inductive generalization of novel properties to same-category or similar-looking objects was studied in Chinese preschool children. The effects of category labels on generalizations were investigated by comparing basic-level labels, superordinate-level labels, and a control phrase applied to three kinds of stimulus materials: colored photographs (Experiment 1), realistic line drawings (Experiment 2), and cartoon-like line drawings (Experiment 3). No significant labeling effects were found for photos and realistic drawings, but there were significant effects for cartoon-like drawings. Children made mostly (>70%) category-based inferences about photographs whether or not labels were provided (Experiment 1). Children showed a bias toward category-based inferences about realistic drawings (Experiment 2) but did so only when labels were provided. Finally, children made mostly appearance-based generalizations for cartoon-like drawings (Experiment 3). However, labels (basic or superordinate level) reduced appearance-based responses. Labeling effects did not depend on having identical labels; however, identical superordinate labels were more effective than different basic-level labels for the least informative stimuli (i.e., cartoons). Thus, labels sometimes confirm the identity of ambiguous items. This evidence of labeling effects in Mandarin-speaking Chinese children extends previous findings beyond English-speaking children and shows that the effects are not narrowly culture and language specific.