German in childhood and Latin in adolescence: On the bidialectal nature of lexical access in English

The study explores the largely unstudied relationship between word etymology and word recognition in English. English, possessing a dual lexical heritage from Germanic and Latin roots, presents a unique case for investigating this relationship. Previous research has extensively examined the Germanic and Latin origins of English words from a linguistic perspective, but the impact of this distinction on word recognition processes has been largely overlooked. The researchers hypothesize that early-acquired words will predominantly originate from Germanic roots, while later-acquired words will be predominantly Latin-based. This hypothesis will be tested by examining word recognition performance (reaction times and accuracy) in relation to word etymology, controlling for factors like AoA, frequency, and length. Additionally, the researchers aim to compare native and non-native English speakers to investigate differences in lexical access patterns. This research contributes to a more comprehensive understanding of word recognition models and lexical access mechanisms in diverse linguistic populations. Understanding these processes has implications for language acquisition and education, particularly for individuals from linguistically diverse backgrounds.

The literature review discusses previous research on age of acquisition (AoA) effects in word recognition across various tasks, including reading, lexical decision, picture naming, and eye-tracking. These studies consistently show that AoA significantly impacts performance, even when controlling for other variables like frequency and length. The network plasticity hypothesis suggests that early-learned words occupy a privileged position in the semantic network, explaining the AoA effect. While initially viewed through a neuroplasticity lens, more integrated views consider multiple frameworks to explain AoA effects, including mapping theory and the role of semantic richness and connections. The study also touches upon research highlighting the differences between Germanic and Latin-based words in terms of their age of acquisition, frequency, and usage in different registers (narrative vs. expository). Previous studies have shown sensitivity to etymology in various linguistic domains, like language production and reading, and suggest a potential bilingual-like processing in monolingual English speakers. However, the impact of etymology on word recognition remains largely unexplored, particularly with a larger and more varied word corpus extending beyond previously investigated AoA ranges.

The study employed a large dataset of approximately 20,000 words drawn from the ELP and supplemented with AoA norms from Kuperman et al. (2012) and etymology information from the Online Etymology Dictionary. Two additional datasets from the ECP (native and L2 English speakers) were included. Word etymology was determined by undergraduate raters, who classified words into Germanic, Latin, or neither origin, considering root, prefix, and suffix. The final database was checked for accuracy. The analyses were conducted in three steps. First, the distribution of Germanic and Latin-based words across different AoA ranges was analyzed using regression and chi-square tests. Second, linear regressions examined the effect of word etymology on reaction times and accuracy in word reading tasks (lexical decision and naming), controlling for AoA, log word frequency, and length. Third, planned comparisons contrast analyses were used to establish differences in accuracy and reaction times based on etymology combinations. The study used data from English monolingual participants (ELP naming task, ELP and ECP lexical decision tasks) and L2 speakers (ECP lexical decision task).

The ELP data for native English speakers showed a strong relationship between AoA and etymology. Germanic single words were acquired significantly earlier than Latin-based words. For compound words, the etymology of the second root mattered only when the first root was Latin; Latin-Latin compounds were acquired later than Latin-Germanic compounds. A chi-square test showed that Germanic words were more prevalent in early childhood, while Latin-based words increased proportionally with age. In the word naming task, native speakers showed faster reaction times and higher accuracy for words with Germanic roots, even after controlling for other variables. Latin-Latin compounds had the slowest reaction times and lowest accuracy. In the lexical decision task, the ELP data showed similar patterns. However, the ECP data for native speakers showed less pronounced effects of etymology on accuracy. In the ECP lexical decision task with L2 speakers, a different pattern emerged. L2 speakers exhibited faster reaction times and higher accuracy for Latin-based words compared to Germanic words. This suggests that L2 learners might prioritize acquiring more advanced, Latin-based vocabulary.

The findings support a bidialectal view of English lexical access, where a Germanic base is established in childhood, and Latin-based vocabulary is integrated later. The results are consistent with previous research on AoA effects and offer insights into the complex interplay between etymology, AoA, and lexical processing. The different patterns observed between native and L2 speakers highlight the varied pathways of language acquisition. The study's findings challenge traditional models of AoA effects, suggesting a more nuanced understanding is needed beyond simple timing of learning. The study emphasizes the importance of considering etymology in future research on word recognition and language acquisition. The observed differences between native and non-native speakers underscore the need for further research to understand the impact of L1 on English acquisition and lexical processing.

This study provides compelling evidence for a bidialectal nature of English lexical access, showing a clear distinction in processing between words of Germanic and Latin origin, particularly in native speakers. This has implications for language acquisition theories and educational practices. Future research could explore the impact of specific L1s on this bidialectal processing, and how these patterns might affect academic achievement, especially for L2 learners.

The study's use of undergraduate native speakers in the ELP dataset might limit the generalizability of the findings to broader populations with diverse linguistic backgrounds and socioeconomic statuses. Future studies should expand the participant pool to include more diverse groups, including individuals with varying levels of English proficiency and different L1s. Further research is needed to fully understand the role of specific L1s and how they interact with the bidialectal structure of English.