Syntactic Error Analysis of Indonesian as a Foreign Language Learners in Social Media Digital Corpora

Abstract

Acquiring syntactic competence is fundamental for Indonesian as a Foreign Language (BIPA) learners, yet structural deviations remain a persistent challenge. While traditional error analysis often relies on controlled classroom data, social media provides a valuable repository of unelicited learner language to observe spontaneous interlanguage development. This study aims to analyze syntactic errors produced by BIPA learners in social media digital corpora and identify patterns of first language interference using the Error Analysis framework. Employing a descriptive qualitative design supported by quantitative statistics, an authentic corpus was purposively sampled from public social media platforms. Data processing involved AI-assisted initial categorization followed by manual researcher verification against standard Indonesian grammar. The results indicate that structural deviations are predominantly rooted in morphological (40%) and mixed (30%) error categories. The most frequent patterns involve phrase sequence inversions and morphosyntactic inaccuracies, particularly the omission or misapplication of mandatory verbal affixes. These anomalies appear driven by negative cross-linguistic transfer from learners' native typologies, alongside intralingual overgeneralization. In conclusion, the study suggests that syntactic competence in Indonesian is closely tied to morphological development. The findings support utilizing authentic digital corpora for language research and highlight the need for data-driven, form-focused BIPA instruction to address specific structural challenges.