Grapheme–Morphonological Difficulties of English-Speaking Learners at the Initial Stage of Learning Kazakh: Toward Justifying an Independent Pre-A1 Level

This study is relevant because the earliest stage of Kazakh as a foreign language remains insufficiently described for English-speaking learners, whose initial acquisition involves not only elementary vocabulary and grammar, but also entry into a new script, unfamiliar sound categories, vowel harmony, and affixal allomorphy. This article aims to justify the need for an independent pre-threshold Pre-A1 stage in Kazakh instruction and to identify the grapheme-phonological and morphonological difficulties that hinder learners’ transition to A1. The study is based on an eight-week intensive A1 Kazakh course at the University of Oxford, involving ten native speakers of English across three consecutive groups, and combines structured pedagogical observation, written assignments, audio-supported phonetic exercises, contrastive analysis of the 42 graphemes of the Kazakh Cyrillic alphabet, and qualitative interpretation of recurrent learner errors. The methodological contribution of the study lies in classifying Kazakh Cyrillic graphemes into familiar, deceptive, and unfamiliar groups for Latin-script users and operationalizing this classification through four coefficients: Visual Familiarity Coefficient, Direct Transfer Coefficient, Misfire Rate, and Interference Risk Coefficient. The results show that visual accessibility does not guarantee successful transfer: only 10 of 42 graphemes allow relatively safe direct transfer, while deceptive graphemes create a high risk of orthographically mediated phonological interference. The classroom evidence further shows that substitutions such as /q/ → /k/, /ʁ/ → /g/, and /ø/ → /y/ affect not only pronunciation, but also the recognition of vowel harmony and suffix selection. The findings substantiate a separate Pre-A1 module organized around script orientation, contrastive sound-letter mapping, perceptual training, vowel-harmony recognition, controlled affixal variation, and embedded correction.