Explicit vs. Implicit Learning of Chinese Numeral Classifiers and Native-English Speakers’ Object Categorization Preferences Yee Pin Tio, M.A., Department of Psychology Email:
[email protected] Usha Lakshmanan, Ph.D., Department of Psychology Introduction
Method
This project sought to address the cognitive effects of numeral classifier exposure by examining whether Native-English speakers’ learning of Chinese numeral classifiers impacts how they categorize objects.
Implicit-Learning vs. Explicit-Learning • Classifier system pose considerable challenges to young Chinese children and Chinese language learner due to : -the complexity and inconsistencies within the classifier system -discrepancy in learning condition and instructional methods (Zhang & Lu, 2012)
• Participants (N=153) consisted of native-English speaking undergraduate students with no prior knowledge of any classifier language (e.g., Korean, Thai and Sign Language), who were randomly assigned to one of the experimental groups (i.e., implicit-learning or explicit-learning) or the control group.
Research Problem: Does explicit or implicit instructions on Chinese numeral classifiers influenced Native-English speakers’ object categorization preferences ?
• Explicit-instructions is operationalized as the presence of rule-explanation as a part of instruction; Implicit-instructions is operationalized as the absence of rule presentation and instruction to attend to particular forms (DeKeyser, 1995).
• The Experimental groups (but not the control group) was systematically exposed to Mandarin Chinese numeral classifiers using a visual, auditory and haptic integration module (see Fig.1)
• Previous studies on Chinese numeral classifiers from a cognitive perspective showed the influence of classifiers on Chinese speakers' performance on memory, attention and inductive reasoning tasks (e.g., Huang & Chen, 2014; Huettig et. al., 2010; Imai, Saalbach & Stern, 2010).
• The positive effects of explicit language learning have been documented in many studies (see Ellis, 2002; Norris & Ortega, 2000). However, implicit knowledge is retained more easily and longer than explicit knowledge (Allen & Reber,1980) . Implicit instruction is more beneficial than explicit instruction in learning fuzzy prototypical morphological rules (DeKeyser,1995).
• However, these studies failed to establish a causal relationship between language and cognitive performance due to lack of random assignment. • A training design controls for language external factors (e.g., culture, age) by comparing, for example, native-English speakers trained in the target language specific structure (e.g., Chinese numeral classifiers) with another (i.e., untrained) group of native-English speakers, who have been randomly assigned to one of the three conditions (i.e., explicitlearning, implicit-learning and control group).
• As Chinese NCs exemplify such rules, we sought to compare the effects of explicit versus implicit approaches to the learning of numeral classifiers on object categorization.
• The control group did not undergo training on Chinese numeral classifiers but completed a modified version of the tasks of approximately the same duration as the training phase. • Forced-Choice Task were used to evaluate the participants’ object categorization preferences. The objects in the task were selected based on an object similarity rating Results completed by an independent group of Native-English speakers. p < 0.01 p < 0.01
Hypotheses: 1. The experimental groups (i.e., explicit-learning & implicit-learning), which was exposed to Chinese numeral classifiers, will demonstrate a bias towards grouping together objects sharing the same classifier. 2. The implicit-learning group would outperform the explicit-learning group by showing stronger preference for classifier-based categorization.
• Differences in performance on cognitive tasks can then be attributed to the language exposure that was manipulated in the study, thus showing the directional influence of language on cognition. • Numeral Classifiers : Grammatical morphemes that occur adjacent to a Table 1 numeral (e.g., one, two) or a determiner (e.g., this, that) and precede the noun in abetween noun phase. Comparison English (non-classifier language) measure words and Mandarin (classifier
language) numeral classifiers.
Figure 3: One-Way ANOVA results comparing experimental and control group means for Classifier-Based Object Selection
Table 1: Numeral Classifiers in Mandarin Chinese
Language
Uncountable Nouns
Countable Nouns
English
One glass of water
One
Mandarin Chinese
一 杯 Yi4 bei1 ‘One CL
一 支 笔 Yi4 zhi1 bi3 ‘One CL pen’
水 shui3 water’
Results & Conclusion
pen
Forced-Choice task • Hypothesis 1 supported. Significant effect of exposure to classifier knowledge on classifier-sharing object categorization, F (1, 151) = 47.65, p < .01, ω= 0.614. •
A planned contrast revealed that having any exposure to Chinese numeral classifiers significantly increased classifier-based categorization compared to the controls, t (151) = 8.637, p = < .01, r =0.233 . Explicit –learning group showed significantly higher classifier-base categorization than implicit-learning group, t (151) = 4.646, p < .01, r = 0.173.
•
Hypothesis 2 not supported. Explicit-learning group selected a significantly (p
