Foreign Accent

Definition:

Foreign accent refers to the systematic set of phonetic and phonological features in an L2 speaker’s speech that deviate from native-speaker norms of the target language—reflecting L1-influenced pronunciation patterns, incomplete acquisition of L2 phonological categories, and sometimes stable non-native patterns that persist despite extensive exposure and instruction. Accent is not a unitary phenomenon but a collection of segmental (individual sounds) and suprasegmental (rhythm, intonation, prosody, pitch) features. Research on foreign accent includes the degree to which L1 phonological transfer accounts for it (Flege’s Speech Learning Model, Best’s Perceptual Assimilation Model), the role of the Critical Period Hypothesis in predicting later-onset greater accent, the intelligibility vs. accentedness distinction, and the sociolinguistic dimensions of accent as identity marker and target of discrimination.

In-Depth Explanation

What produces foreign accent?

Foreign accent arises from multiple sources:

L1 phonological transfer: Learners map L2 sounds onto L1 phonological categories — segments without L1 equivalents are approximated by the nearest L1 sound, and L1 prosodic patterns (stress, rhythm, intonation) are superimposed onto the L2 phonological system.
Perception before production failure: Learners cannot produce contrasts they cannot perceive. Cross-linguistic phonetic perception research (Flege 1995; Best 1995) shows that L2 sounds similar to L1 sounds are actually harder to perceive accurately than sounds very different from L1 — similar sounds are assimilated to the L1 category (Perceptual Assimilation Model).
Suprasegmental mismatch: Stress patterns, rhythm, intonation contours, and timing (mora-timed vs. stress-timed vs. syllable-timed) differ substantially across languages. Japanese is mora-timed; English is stress-timed — learners of each direction carry rhythmic patterns from their L1.
Critical period effects: The Critical Period Hypothesis (Lenneberg 1967; Scovel 1988; Long 1990) predicts that age of acquisition before approximately 6–12 increases the probability of native-like phonological acquisition; after this period, accent is increasingly likely. Empirical evidence supports a phonological sensitive period, though not a hard cutoff.

Flege’s Speech Learning Model (SLM):

Flege (1995) proposed the Speech Learning Model as an alternative to L1-interference-only accounts:

The phonological system remains plastic throughout life — adult L2 learners can form new phonetic categories.
New L2 categories are most likely to form for sounds that are dissimilar from all L1 sounds — sounds very different from L1 are easier to acquire as distinct categories.
L2 sounds that are similar to L1 sounds are hard to separate from L1 categories — they are perceptually assimilated, and a new category is rarely formed.
This predicts a similarity paradox: Similar L2 sounds cause more persistent accent than highly dissimilar sounds (which more easily trigger new category formation).
SLM allows for lifelong plasticity but predicts increasing difficulty with age due to L1-L2 category interference.

Best’s Perceptual Assimilation Model (PAM):

Best (1995) proposed that non-native sound perception depends on how well non-native sounds can be assimilated to L1 phonetic categories:

Two-category assimilation: One non-native contrast maps onto two L1 categories — easiest to discriminate.
Single-category assimilation: Both phones in a contrast map to the same L1 category — hardest to discriminate.
Category goodness difference: Both map to same L1 category but one is a better exemplar — intermediate difficulty.
PAM predicts which L2 contrasts will be easy vs. difficult for a given L1 background.

Accentedness vs. intelligibility:

Derwing and Munro’s research (1997, 2009) introduced a crucial distinction:

Accentedness: The degree to which a listener perceives a speaker as non-native — subjective judgment, sensitive to listener L2 contact, attitudes, and pronunciation rating.
Intelligibility: The degree to which a listener can understand what was said — communicatively more important.
Comprehensibility: The degree to which the listener finds understanding effortful.

These three dimensions do not always correlate: a heavily accented speaker may be highly intelligible (strong accent but all words understood); a mildly accented speaker may be hard to understand (subtle deviations at suprasegmental level reduce comprehensibility). Pronunciation instruction should target intelligibility and comprehensibility rather than accent reduction per se.

Japanese L2 English accent:

Japanese L1 learners of English show characteristic foreign accent features:

Vowel epenthesis: English consonant clusters (e.g., strike → sutoraikl) are broken up by inserted vowels following Japanese CV syllable structure.
L/R confusion: The Japanese /r/ (a flap, [ɾ]) is a distinct phonetic category from English /l/ and /r/; English /l/ and /r/ may be perceptually assimilated to a single Japanese category.
Length distinction not phonemically important in English: Japanese phonemically contrasts vowel length (お vs. おお; kite vs. kiite); this distinction is not phonemically relevant in English, but transfer in the other direction (English learners of Japanese) produces length errors.
Pitch accent → stress transfer: Japanese is a pitch-accent language without lexical stress; Japanese L1 learners of English may not consistently produce English stress-timed rhythm.

English L1 learner accent in Japanese:

English L1 learners of Japanese show characteristic features:

Mora timing: Japanese is mora-timed (each mora, including N and long vowels, takes equal time); English speakers produce syllable-timed Japanese, distorting prosodic patterns.
Pitch accent errors: Tokyo dialect has a pitch accent system (H/L tone assignment by accent class) that English speakers must acquire; without explicit instruction, English speakers default to stress (dynamic) rather than pitch (tonal) prominence.
Geminate consonants: Japanese geminates (kitte vs. kite) require a distinctive consonant duration that English speakers must acquire — English has no productive geminate system.
N moraic conservation: The syllabic/moraic nasal /N/ (ん) must be produced as a separate mora — English speakers tend to collapse it into the preceding vowel.

History

1967: Lenneberg — Critical Period Hypothesis; age effects on accent.
1985: Major — Ontogeny Model relating L1 transfer and developmental sequences in L2 phonology.
1988: Scovel — accent as CPH evidence.
1990: Long — critical period for L2 phonology review.
1995: Flege — Speech Learning Model.
1995: Best — Perceptual Assimilation Model.
1997: Derwing & Munro — accentedness vs. intelligibility distinction.
2005–present: Explicit pronunciation instruction research; Critical Period debate continues.

Common Misconceptions

“Native-like accent is impossible for adult learners.” Research on late learners documents a small minority who achieve near-native pronunciation — the CPH predicts greater difficulty, not impossibility. Explicit pronunciation instruction, high phonetic aptitude, and extended high-quality input can produce very low accent in adult learners.

“Foreign accent indicates low proficiency.” Accent and proficiency are distinct dimensions. A speaker may have near-native grammar and vocabulary but strong accent; and a speaker with a mild accent may have limited grammar. Conflating accent with proficiency is a social bias, not a linguistic reality.

“Accent reduction is the goal of pronunciation instruction.” Contemporary pronunciation pedagogy targets intelligibility and comprehensibility as communicatively relevant goals — accent reduction for cosmetic purposes conflates social normativity with communicative competence.

Criticisms

The CPH for phonology has been critiqued — defining the critical period boundaries (6? 12? puberty?) is inconsistent across studies, and counter-examples of late-onset near-native phonology complicate strict biological interpretations.
SLM has been criticized for complexity and for not providing precise predictions about all L2 phonological acquisition challenges.
The accentedness/intelligibility distinction, while important, is difficult to apply pedagogically — some accent features affect intelligibility, others do not, and which features to prioritize for instruction requires empirical guidance.

Social Media Sentiment

Foreign accent in Japanese is a highly discussed topic among L2 Japanese learners. Many advanced learners express pride in low-accent pronunciation achieved through deliberate shadow reading, pronunciation study, and native-speaker audiovisual input. Others note that Japanese native speakers are often very forgiving of English-accented Japanese (possibly more so than in English-speaking contexts with foreign-accented English). Pitch accent study (Tokyo dialect accent class learning) has become a significant component of advanced Japanese phonological study in online learning communities.

Last updated: 2026-04

Practical Application

Target prosody, not just segments: For Japanese, mora timing, geminate length, long vowel duration, and pitch accent are the suprasegmental features that most affect comprehensibility. Individual vowel/consonant errors matter less.
Explicit pitch accent study: Study Tokyo dialect accent class patterns systematically — resources like the Dogen pitch accent course and Suzuki-kun pitch accent dictionaries provide explicit instruction for pitch accent acquisition.
Shadowing methodology: Closely time-synced pronunciation practice with native audio (shadowing) activates the phonological feedback loop and allows learners to compare their production to native models in real time — more effective than production practice without models.
Perception training before production: Systematic minimal pair listening practice (L Japanese: kite vs. kiite, kami with accent class 2 vs. accent class 1) trains the perceptual distinctions that production depends on.

Related Terms

Research

Flege, J. E. (1995). Second language speech learning: Theory, findings, and problems. In W. Strange (Ed.), Speech Perception and Linguistic Experience: Issues in Cross-Language Research (pp. 233–277). York Press. [Summary: Speech Learning Model; L1-L2 phonetic category interaction; similarity paradox; lifelong phonological plasticity with L1 interference; most influential L2 phonological acquisition model.]

Best, C. T. (1995). A direct realist view of cross-language speech perception. In W. Strange (Ed.), Speech Perception and Linguistic Experience (pp. 171–204). York Press. [Summary: Perceptual Assimilation Model; how non-native sounds are assimilated to L1 categories; predictions for discrimination difficulty; phonetic perception framework foundational for L2 sound acquisition research.]

Derwing, T. M., & Munro, M. J. (1997). Accent, intelligibility, and comprehensibility: Evidence from four L1s. Studies in Second Language Acquisition, 19(1), 1–16. [Summary: Accentedness vs. intelligibility distinction established; measurement methodology; different L1 backgrounds differ in intelligibility impact; foundational for pronunciation pedagogy orientation.]

Lenneberg, E. H. (1967). Biological Foundations of Language. Wiley. [Summary: Critical Period Hypothesis; language acquisition on genetic/biological schedule; puberty as end of sensitive period; foundational for CPH debate applied to L2 phonological acquisition.]

Long, M. H. (1990). Maturational constraints on language development. Studies in Second Language Acquisition, 12(3), 251–285. [Summary: Review of CPH evidence for L2 acquisition; critical period for phonology most supported; less clear for morphosyntax; influential review connecting CPH to L2 research.]