A cross-linguistic investigation into children’s acquisition of the successor function

• In Event: 1-030 - Linguistic basis of numerical cognition

Integrative Statement

This study explores the hypothesis (Miller & Stigler, 1987) that children who speak a language in which the base system of number is transparent are faster to become competent counters, relative to children learning a less transparent system. On this account, the transparency of the base system across languages is causally related to the acquisition of productive counting knowledge, which in turn is related to learning other mathematical knowledge. Although previous research has provided preliminary evidence for this hypothesis, these studies have reported between-group differences which may be explained by alternative factors, such as differences in training amount or cultural differences in educational techniques. Here, we addressed this idea by asking how counting productivity differs across languages that differ in transparency, and how this is related to learning the successor function – i.e., that every number has a successor defined as N+1.
First, we compared children both across cultures (English, Cantonese, and Slovenian) and within culture (Hindi vs. Gujarati). The comparison of Slovenian and Cantonese provides a critical test of the hypothesis that linguistic transparency drives productive counting knowledge, as these two groups are known to have relatively transparent counting systems, but very different levels of exposure to counting training in early childhood. Meanwhile, the counting systems of Gujarati and Hindi, both spoken in the Indian state of Gujarat, are etymologically related but differ in transparency (with Hindi being much less transparent than many other languages). Second, in addition to group-wise comparisons, we planned within-group comparisons to test the relation between productive counting ability and knowledge of the successor function within each language group.
We tested 500+ children between the ages of 3;6 and 6;6. Children completed five tasks: Give-N, Highest Count (with prompts given at errors), the Unit/Successor Task (Sarnecka & Carey, 2008), the What Comes Next Task, and a picture memory task. We find some preliminary evidence that, while counting proficiency differs between language groups, mastery of the base system is related to performance on the Unit and What Comes Next tasks. This effect is especially pronounced in languages with lower exposure to the count routine. Thus, while we find fewer children with productive knowledge of the base system in languages with less counting exposure (like Slovenian), nevertheless “Productive” counters perform similarly across languages that differ in transparency.
Taken together, this large dataset consisting of 5 languages will allow us to test the hypothesis that languages with more transparent counting systems facilitate productive counting knowledge in children either earlier in life, or with less training (relative to children speaking less transparent languages).

Authors

• Rose Schneider, University of California San Diego
  Presenting Author

• Jess Sullivan, Skidmore College
  Non-Presenting Author

• Kaiqi Guo, University of California San Diego
  Non-Presenting Author

• Priyanka Biswas, University of California San Diego
  Non-Presenting Author

• David Barner, University of California, San Diego
  Non-Presenting Author