Reproducibility and a unifying explanation: Lessons from the shape bias

Late Talkers can generalise trained labels by object shape similarities, but not unfamiliar labels

Late talkers (LTs) exhibit delayed vocabulary development, which might stem from a lack of a typical word learning strategy to generalise object labels by shape, called the 'shape bias'. We investigated whether LTs can acquire a shape bias and whether this accelerates vocabulary learning. Fourteen LTs were randomly allocated to either a shape training group (Mage = 2.76 years, 6 males), which was taught that objects similar in shape have the same name, or a control group (Mage = 2.61 years, 4 males), which was taught real words without any focus on object shape. After seven training sessions, children in the shape training group generalised trained labels by shape (d = 1.28), but not unfamiliar labels. Children in the control group extended all labels randomly. Training did not affect expressive vocabulary.

Capturing the Heterogeneity of Word Learners by Analyzing Persons

Accurately capturing children's word learning abilities is critical for advancing our understanding of language development. Researchers have demonstrated that utilizing more complex statistical methods, such as mixed-effects regression and hierarchical linear modeling, can lead to a more complete understanding of the variability observed within children's word learning abilities. In the current paper, we demonstrate how a person-centered approach to data analysis can provide additional insights into the heterogeneity of word learning ability among children while also aiding researchers' efforts to draw individual-level conclusions. Using previously published data with 32 typically developing and 32 late-talking infants who completed a novel noun generalization (NNG) task to assess word learning biases (i.e., shape and material biases), we compare this person-centered method to three traditional statistical approaches: (1) a t-test against chance, (2) an analysis of variance (ANOVA), and (3) a mixed-effects regression. With each comparison, we present a novel question raised by the person-centered approach and show how results from the corresponding analyses can lead to greater nuance in our understanding of children's word learning capabilities. Person-centered methods, then, are shown to be valuable tools that should be added to the growing body of sophisticated statistical procedures used by modern researchers.

What curves are parallel? The core feature of preschoolers' intuitive parallel category

Existing evidence has revealed that humans can spontaneously categorize many geometric shapes without formal education. Children around 4 years could distinguish between intersecting lines and parallel lines. Three features can be used to identify parallel lines, namely "translational congruence," "never meet," and "constant distance." This study separated them by using pairs of curves that possess only one of these features. Two experiments across 2021-2023, respectively, compared the relative priority of "translational congruence" with "constant distance," and "never meet" with "constant distance" among 3- to 5-year-old Chinese preschoolers (Ntotal = 314, 48% female). The results showed that preschoolers consistently grouped "constant distance" curves with parallel lines. This suggests that the core feature of intuitive parallel category is "constant distance" at this age.

How children generalize novel nouns: An eye-tracking analysis of their generalization strategies

Recent research has shown that comparisons of multiple learning stimuli which are associated with the same novel noun favor taxonomic generalization of this noun. These findings contrast with single-stimulus learning in which children follow so-called lexical biases. However, little is known about the underlying search strategies. The present experiment provides an eye-tracking analysis of search strategies during novel word learning in a comparison design. We manipulated both the conceptual distance between the two learning items, i.e., children saw examples which were associated with a noun (e.g., the two learning items were either two bracelets in a "close" comparison condition or a bracelet and a watch in a "far" comparison condition), and the conceptual distance between the learning items and the taxonomically related items in the generalization options (e.g., the taxonomic generalization answer; a pendant, a near generalization item; versus a bow tie, a distant generalization item). We tested 5-, 6- and 8-year-old children's taxonomic (versus perceptual and thematic) generalization of novel names for objects. The search patterns showed that participants first focused on the learning items and then compared them with each of the possible choices. They also spent less time comparing the various options with one another; this search profile remained stable across age groups. Data also revealed that early comparisons, (i.e., reflecting alignment strategies) predicted generalization performance. We discuss four search strategies as well as the effect of age and conceptual distance on these strategies.

Object Shape and Depth of Word Representations in Preschoolers

This study examined the effect of a shape cue (i.e., co-speech gesture) on word depth. We taught 23 preschoolers (M = 3;5 years, SD = 5.82) novel objects with either shape (SHP) or indicator (IND) gestures. SHP gestures mimicked object form, but IND gestures were not semantically related to the object (e.g., an upward-facing palm, extended toward the object). Each object had a unique IND or SHP gesture. Outcome measures reflected richer semantic and phonological learning in the SHP than in the IND condition. In the SHP condition, preschoolers (a) expressed more semantic knowledge, (b) said more sounds in names, and (c) generalized more names to untaught objects. There were also fewer disruptions to prime picture names in the SHP condition; we discuss the benefit of a co-speech shape gesture to capitalize on well-established statistical word learning patterns.

For human-like models, train on human-like tasks

Bowers et al. express skepticism about deep neural networks (DNNs) as models of human vision due to DNNs' failures to account for results from psychological research. We argue that to fairly assess DNNs, we must first train them on more human-like tasks which we hypothesize will induce more human-like behaviors and representations.

Vocabulary and automatic attention: The relation between novel words and gaze dynamics in noun generalization

Words direct visual attention in infants, children, and adults, presumably by activating representations of referents that then direct attention to matching stimuli in the visual scene. Novel, unknown, words have also been shown to direct attention, likely via the activation of more general representations of naming events. To examine the critical issue of how novel words and visual attention interact to support word learning we coded frame-by-frame the gaze of 17- to 31-month-old children (n = 66, 38 females) while generalizing novel nouns. We replicate prior findings of more attention to shape when generalizing novel nouns, and a relation to vocabulary development. However, we also find that following a naming event, children who produce fewer nouns take longer to look at the objects they eventually select and make more transitions between objects before making a generalization decision. Children who produce more nouns look to the objects they eventually select more quickly following the naming event and make fewer looking transitions. We discuss these findings in the context of prior proposals regarding children's few-shot category learning, and a developmental cascade of multiple perceptual, cognitive, and word-learning processes that may operate in cases of both typical development and language delay. RESEARCH HIGHLIGHTS: Examined how novel words guide visual attention by coding frame-by-frame where children look when asked to generalize novel names. Gaze patterns differed with vocabulary size: children with smaller vocabularies attended to generalization targets more slowly and did more comparison than those with larger vocabularies. Demonstrates a relationship between vocabulary size and attention to object properties during naming. This work has implications for looking-based tests of early cognition, and our understanding of children's few-shot category learning.

Children's generalization of novel names in comparison settings: The role of semantic distance during learning and at test

There is considerable evidence showing that, in novel noun learning and generalization tasks, comparisons of several learning stimuli lead to more taxonomically based generalizations of novel nouns than single stimulus presentations. This study investigated the role of semantic distance (close vs. far) between learning examples and between learning examples and transfer items (near vs. distant) in comparison designs. In two experiments, we investigated the case of object nouns (e.g., foods, Experiment 1) and relational nouns (e.g., is the cutter for, Experiment 2) in 4- to 6-year-old and 3- to 4-year-old children, respectively. As predicted, the comparison conditions led to better results than the no-comparison conditions. In comparison conditions, far training items and near generalization items gave the best performance. Semantic distance effects are discussed in terms of abstracted representations during learning as well as in terms of cognitive constraints on generalization. It is argued that both object nouns and relational nouns are construed in the light of the type of example used during learning (i.e., single or multiple). Depending on the distance between learning and generalization items, children build different categories and are more or less likely to accept distant referents.

Over-reliance on English hinders cognitive science

English is the dominant language in the study of human cognition and behavior: the individuals studied by cognitive scientists, as well as most of the scientists themselves, are frequently English speakers. However, English differs from other languages in ways that have consequences for the whole of the cognitive sciences, reaching far beyond the study of language itself. Here, we review an emerging body of evidence that highlights how the particular characteristics of English and the linguistic habits of English speakers bias the field by both warping research programs (e.g., overemphasizing features and mechanisms present in English over others) and overgeneralizing observations from English speakers' behaviors, brains, and cognition to our entire species. We propose mitigating strategies that could help avoid some of these pitfalls.

High-Level Visual Encoding Model Framework with Hierarchical Ventral Stream-Optimized Neural Networks

Visual encoding models based on deep neural networks (DNN) show good performance in predicting brain activity in low-level visual areas. However, due to the amount of neural data limitation, DNN-based visual encoding models are difficult to fit for high-level visual areas, resulting in insufficient encoding performance. The ventral stream suggests that higher visual areas receive information from lower visual areas, which is not fully reflected in the current encoding models. In the present study, we propose a novel visual encoding model framework which uses the hierarchy of representations in the ventral stream to improve the model’s performance in high-level visual areas. Under the framework, we propose two categories of hierarchical encoding models from the voxel and the feature perspectives to realize the hierarchical representations. From the voxel perspective, we first constructed an encoding model for the low-level visual area (V1 or V2) and extracted the voxel space predicted by the model. Then we use the extracted voxel space of the low-level visual area to predict the voxel space of the high-level visual area (V4 or LO) via constructing a voxel-to-voxel model. From the feature perspective, the feature space of the first model is extracted to predict the voxel space of the high-level visual area. The experimental results show that two categories of hierarchical encoding models effectively improve the encoding performance in V4 and LO. In addition, the proportion of the best-encoded voxels for different models in V4 and LO show that our proposed models have obvious advantages in prediction accuracy. We find that the hierarchy of representations in the ventral stream has a positive effect on improving the performance of the existing model in high-level visual areas.

The signature-testing approach to mapping biological and artificial intelligences

Making inferences from behaviour to cognition is problematic due to a many-to-one mapping problem, in which any one behaviour can be generated by multiple possible cognitive processes. Attempts to cross this inferential gap when comparing human intelligence to that of animals or machines can generate great debate. Here, we discuss the challenges of making comparisons using 'success-testing' approaches and call attention to an alternate experimental framework, the 'signature-testing' approach. Signature testing places the search for information-processing errors, biases, and other patterns centre stage, rather than focussing predominantly on problem-solving success. We highlight current research on both biological and artificial intelligence that fits within this framework and is creating proactive research programs that make strong inferences about the similarities and differences between the content of human, animal, and machine minds.

A Transformer-Based Capsule Network for 3D Part–Whole Relationship Learning

Learning the relationship between the part and whole of an object, such as humans recognizing objects, is a challenging task. In this paper, we specifically design a novel neural network to explore the local-to-global cognition of 3D models and the aggregation of structural contextual features in 3D space, inspired by the recent success of Transformer in natural language processing (NLP) and impressive strides in image analysis tasks such as image classification and object detection. We build a 3D shape Transformer based on local shape representation, which provides relation learning between local patches on 3D mesh models. Similar to token (word) states in NLP, we propose local shape tokens to encode local geometric information. On this basis, we design a shape-Transformer-based capsule routing algorithm. By applying an iterative capsule routing algorithm, local shape information can be further aggregated into high-level capsules containing deeper contextual information so as to realize the cognition from the local to the whole. We performed classification tasks on the deformable 3D object data sets SHREC10 and SHREC15 and the large data set ModelNet40, and obtained profound results, which shows that our model has excellent performance in complex 3D model recognition and big data feature learning.

Preserved category-based inferences for word learning in school-aged children with developmental language disorder

Word learning difficulties are often found in children with Developmental Language Disorder (DLD). Lexical patterns of difficulties appear to be well described in the context of DLD but very little research focuses on their underlying causes. Word learning is known to be an inference-based process, constrained by categorization, which helps the extension of new words to unfamiliar referents and situations. These processes appear integrated in Bayesian models of cognition, which supposes that learning relies on an inductive inference process that recruits prior knowledge and principles of statistical learning (detection of regularities). Taken together, these mechanisms remain underexplored in DLD. Our study aims to define whether children with DLD can draw inductive inferences in a word learning context using categorization. Twenty children with DLD (between 6;0 and 12;6), and 20 language-matched and 16 age-matched controls were exposed to a word learning task where they were given exemplars of objects associated with pseudo-words. The objects belonged to six categories spread across three hierarchical levels. For each item, the children chose which one(s), among a set of test objects from the same categories, could be labelled the same way (word extension). Results showed that school-aged children with DLD could extend new words to broader categories as well as their typically developing (TD) peers. Nevertheless, none of the DLD or TD children showed a specification of their categorization of familiar instances that referred to more restricted instances. Our study suggests preserved abilities in using conceptual knowledge in order to learn new words, which could be used as a compensative strategy in the context of therapy. Further studies are needed to investigate this ability in more complex learning contexts.

Atypical shape bias and categorization in autism: Evidence from children and computational simulations

The shape bias, a preference for mapping new word labels onto the shape rather than the color or texture of referents, has been postulated as a word-learning mechanism. Previous research has shown deficits in the shape bias in children with autism even though they acquire sizeable lexicons. While previous explanations have suggested the atypical use of color for label extension in individuals with autism, we hypothesize an atypical mapping of novel labels to novel objects, regardless of the physical properties of the objects. In Experiment 1, we demonstrate this phenomenon in some individuals with autism, but the novelty of objects only partially explains their lack of shape bias. In a second experiment, we present a computational model that provides a developmental account of the shape bias in typically developing children and in those with autism. This model is based on theories of neurological dysfunctions in autism, and it integrates theoretical and empirical findings in the literature of categorization, word learning, and the shape bias. The model replicates the pattern of results of our first experiment and shows how individuals with autism are more likely to categorize experimental objects together on the basis of their novelty. It also provides insights into possible mechanisms by which children with autism learn new words, and why their word referents may be idiosyncratic. Our model highlights a developmental approach to autism that emphasizes deficient representations of categories underlying an impaired shape bias.