A learner's first language predicts L2 phoneme perception more strongly than their proficiency level does.¶

Claim · All claims
Evidence · 1 study · associational r? · q2 · n=104
Currency · newest study 2026 · checked against its sources 2026-10-20 (automated, 1 of 1 entry) · not reviewed by a person

The contrast being perceived is the same; the task is not. A learner whose first language already encodes the distinction hears it, and one whose first language does not may not hear it at any level.

Subclaims¶

q2 i? Identification accuracy on the Italian singleton–geminate contrast ordered by first language — Italian controls 96.2%, Finnish 89.8%, Czech 83.3%, Slovak 83.1%, Spanish 75.1%, German 64% — with learners whose L1 encodes phonological quantity performing better. → Pešková 2026

q2 i? Accuracy tended to rise with proficiency, but that effect varied across L1 groups, was not significant and was less systematic than the influence of first-language background. → Pešková 2026

q2 i? Identification accuracy (81% overall, with orthographic support) exceeded the same participants' accuracy in an earlier discrimination study in every group except the Finnish learners, which the author takes to suggest that explicit categorical decisions with written options help where fine-grained acoustic sensitivity alone does not. → Pešková 2026

Evidence¶

Pešková 2026¶

Pešková, A. (2026). Identification of consonant length in L2 Italian: the role of cross-linguistic influence. Frontiers in Language Sciences, 5. doi:10.3389/flang.2026.1911194

q2 · cross-sectional group comparison with a native-speaker control group · i? · accuracy percentages by group (89.8% vs 64% on the same contrast); no standardized effect size printed · n=104 · associational · r?

104 learners of Italian identified singleton against geminate consonants in trisyllabic pseudowords (sàpolo vs sàppolo), grouped by first language. Accuracy tracked whether the learner's L1 encodes phonological quantity in consonants or vowels, L1 background had the stronger and more systematic effect; accuracy tended to rise with proficiency, but that effect varied across L1 groups and was not significant. The experiment was an identification task (with orthographic support); compared with a previous discrimination study with the same participants, identification was more accurate in every group except the Finnish learners.

Discussion¶

A difficulty ordering derived from proficiency alone will be wrong for a mixed-L1 cohort. The same item is a different task for a Finnish learner and a German one, and adaptive sequencing that models only level cannot see the difference. This is the perception-side counterpart of a more general point: Prior Knowledge Determines New Learning [+S], where the relevant prior knowledge is a phonological inventory rather than domain content.

It separates two failures that look identical in the data. A learner who does not know a contrast and a learner who cannot hear it both score low. The second will show as weak performance in listening and normal performance in writing, and treating them as the same evidence understates what the learner knows.

Scope. One contrast in one language, and a perception task rather than production. Whether the effect size holds for grammatical rather than phonological contrasts is untested here — though Italian grammatical gender stays incomplete for L2 and heritage speakers alike [~M] finds a related conditionality in morphology, where a typologically closer L1 helps only in some configurations.

Learner Variables¶

• English-Learner Status — predictor: learners who differ on it differ in outcomes

Related Claims¶

• Italian grammatical gender stays incomplete for L2 and heritage speakers alike — the same "L1 helps, but conditionally" shape, in morphology rather than phonology
• Game-based practice outperforms traditional L2 vocabulary instruction — carries the parallel finding that Italian articles are disproportionately hard for learners whose L1 has no article system
• Fluent bilingualism is linked to metalinguistic awareness of how the two languages relate — the other direction: what a learner's language background gives them rather than costs them
• Phoneme awareness is a stronger predictor of reading development than rhyme awareness — phonemic sensitivity as a predictor, in first-language reading
• Prior Knowledge Determines New Learning — the general principle this instantiates
• Preservation of the mother tongue predicts second-language achievement among migrant children — related
• Morpheme difficulty orders were similar across first-language backgrounds, suggesting universals in language acquisition — related
• Adults may have superior language learning capabilities in vocabulary and language structure — related
• A learner's mother language affects second language acquisition, particularly in production of absent sounds — related
• Efficiency in phonological and orthographic processing contributes significantly to individual differences in adult ESL reading measures — related
• Cross-linguistic transfer is skill-specific: phonological awareness and print concepts are more likely to transfer across languages, while vocabulary and comprehension are more language-dependent — related
• The phonological loop plays a significant role in vocabulary acquisition and second language learning — related
• Ultimate attainment and optimal-period length do not reliably differ across major native-language groups, with slightly more evidence for differences in learning curves — related
• Young DLLs have two separate phonological and grammatical language systems early in life