Abstract
This study investigated the concrete modality of network-type thinking and knowledge entity formation of young children aged from four to six, employing picture books. More precisely, we analyzed the techniques with which they associate words and concepts from read stories. As a result, it was demonstrated that the children substantially hire verbs and syntagmatic association when developing a knowledge network. In addition, it was also shown that six-year-olds recalled significantly more keywords than younger children when it came to the syntagmatic (or “serial”) thinking method. Furthermore, this paper indicates some of the future possibilities of the findings that are to be applied to the more effective selection of picture books in nursery school settings for their language development.
Introduction: The Architecture of Early Knowledge
Human wisdom is fundamentally structured as a network—a web of interconnected nodes rather than a collection of isolated facts. While adults typically organize these networks through paradigmatic (hierarchical or categorical) relationships—for instance, recognizing that a cat and a dog both belong to the category of 'animals'—the cognitive processes of young children are often more intuitive and narrative-driven. This research seeks to demystify the "modality" of children's thinking by identifying the "edges" that connect the "nodes" of their knowledge during early development. Understanding these connections is a cornerstone of "eduinformatics," a field that merges education with informatics to optimize learning materials.
Methodology
The study involved 119 children from Japanese kindergartens, categorized into three age groups: 4-year-olds, 5-year-olds, and 6-year-olds. The researchers employed a "storybook connection game" consisting of four phases:
- Recitation: Two picture books, Otsukai (Errands) and Thomas tidies his room, were read to the participants.
- Recall: Children were given two minutes to list keywords remembered from each story.
- Association: Participants selected one keyword from each book and verbalized a connection (e.g., linking a bear from one book to a raincoat from another).
- Analysis: Text mining and word cloud visualizations were used to identify patterns, followed by an Analysis of Variance (ANOVA) to determine statistical significance across age groups.
Key Insights and Ideas
The analysis revealed that children do not merely see objects; they see potential actions. While the recalled keywords were primarily nouns, the associations between them were dominated by verbs such as "play," "put," "eat," and "wear." These verbs act as the "glue" that binds disparate concepts into a coherent network.
Furthermore, children showed a distinct preference for syntagmatic (serial) thinking over taxonomy. Instead of grouping items by category, they linked them through narrative sequences—for example, connecting a cat and a mouse through the action of a chase rather than the category of "animals." Younger children also frequently employed animistic thinking, anthropomorphizing objects to create emotional or social links.
Key Results
The data indicates a significant developmental shift, often referred to as the "Six-Year-Old Leap." Six-year-olds demonstrated a superior ability to recall keywords (averaging 8.5 compared to 4.8 for 4-year-olds) and formed significantly more associations (5.5 versus 2.1).
Quantitatively, the mean number of serial (narrative) connections for 6-year-olds was 4.1, while parallel (categorical) connections remained low at 1.2. The increase in serial thinking was statistically significant ($p < 0.01$), whereas the growth in parallel thinking was negligible. This suggests that abstract classification and paradigmatic thinking likely emerge more fully after children enter primary school.
Limitations and Future Directions
The researchers noted a gender imbalance in the sample, with a higher proportion of female participants. Given that girls often show earlier neural connectivity in language-processing regions, future studies should aim for a more balanced demographic and use "month age" for higher granularity.
The findings suggest significant potential for the development of AI-assisted systems (using Large Language Models) to help educators select picture books that specifically trigger and strengthen these emerging narrative network patterns, ultimately enhancing early childhood language development.
Summary updated