Collocational Errors of Learners of Kazakh as a Second Language: Based on the National Corpus Data

Nazira Amirzhanova, А. Фазылжан

2026

Full text Open access: Hybrid Metadata (CSL-JSON)
DOI 10.55491/2411-6076-2026-3-121-132
Data confidence 85 high
How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

The written production of learners of Kazakh as a second language (L2) has not yet been examined with corpus-based methods. The aim of this study is to identify and classify collocational errors in the written texts of learners of Kazakh as a second language and to interpret their causes using learner corpus data. This study focuses on identifying, classifying and interpreting collocational errors in a learner corpus. It contributes to second language acquisition (SLA) theory with evidence from agglutinative Turkic languages, Kazakh in particular. Building on previous research, the article proposes the theoretical category of “cultural collocation”. The data consist of a corpus of 65 essays on the topic “The Nauryz Holiday” written by 25 secondary school students (12,500-14,000 words in total). The essays were annotated manually and analysed using the Error Analysis approach. A total of 94 collocational errors were identified and classified into four categories: lexical substitution (47%), deviations in verb-noun combinations (24%), errors related to first language influence (18%) and errors in adjective-noun combinations (11%). The analysis shows that grammatical accuracy and collocational competence diverge. The findings provide an empirical basis for improving L2 Kazakh curricula, lexicographic resources and teacher training.