Why Phone Calls in a Second Language Feel Unbearable

You can handle a face-to-face conversation in German. Maybe not elegantly, but you get through it. Then the phone rings — a doctor's office, a landlord, the Krankenkasse — and your mind goes blank before you even answer.

This is not a character flaw. It is a predictable response to a communication medium that strips away the very compensatory mechanisms L2 speakers depend on. Phone calls in a second language are harder. The research confirms what your nervous system already knows.

When you listen to your second language, your eyes do work your ears cannot. Eye-tracking studies consistently show that L2 speakers look significantly more at the speaker's mouth than native speakers do — and lower-proficiency listeners fixate even more. The phone removes precisely this compensation channel.

Birules et al. (2020) found that even highly proficient L2 speakers attended significantly more to the talker's mouth when processing L2 versus their native language. Gruter et al. (2023) replicated and extended this: lower-proficiency listeners showed a gradient effect, fixating even more on the mouth. This means phone calls eliminate the visual channel that L2 speakers actively exploit.

Cognitive Load Theory explains why phone calls feel overwhelming. Working memory has limited capacity distributed across partially independent visual/spatial and auditory/verbal channels. L2 listening already imposes high intrinsic load from unfamiliar phonology, syntax, and vocabulary. When visual cues are available, they provide a compensatory channel. On the phone, all processing concentrates on the auditory channel.

The result is a resource squeeze: your brain needs more processing power for the same input. When comprehension difficulty rises, the same objective speech rate feels subjectively faster because you cannot segment and predict as effectively. The literature consistently identifies speech rate as a major L2 listening difficulty — and on the phone, this is amplified.

Repair — the process of fixing misunderstandings — becomes a double bind on the phone. Breakdowns are more frequent because you cannot lip-read or use gestures. Yet the available repair strategies are more limited and more face-threatening. All repair must be accomplished verbally. Studies of L2 phone conversations find a pattern of repair avoidance: service personnel accept candidate understandings rather than initiating potentially face-threatening repair sequences.

Perceived control is a transdiagnostic anxiety vulnerability — it cuts across social phobia, generalized anxiety, panic, and OCD. A meta-analysis of 51 studies (N=11,218) found a large negative association between perceived control and anxiety. L2 phone calls embody reduced perceived control: you cannot predict the interlocutor's vocabulary, speaking rate, accent, or topic shifts.

Script theory explains why unpredictable calls feel worse. Mental representations of stereotyped event sequences — the 'restaurant script,' the 'phone call script' — enable prediction and reduce cognitive processing demands. When you lack well-developed target-language phone scripts, the predictability advantage is lost. Mak (2011) found that 'speaking without preparation' was the most anxiety-provoking factor among 313 Chinese ESL students.

Mowrer's two-factor theory explains the reinforcement cycle precisely. Factor one: classical conditioning pairs the phone with aversive experiences (embarrassment, incomprehension), creating a conditioned fear response. Factor two: avoiding the call produces immediate anxiety relief (negative reinforcement), strengthening avoidance behavior.

Critically, low-cost avoidance behaviors are resistant to fear extinction — even after the fear response objectively decreases, if the avoidance option remains available, people revert to it. Phone call avoidance is precisely this type of low-cost avoidance: texting, emailing, or asking a partner to call are easy substitutes that feel like reasonable alternatives.

The dependency cycle undermines the practice that builds proficiency. Swain's Output Hypothesis holds that L2 speakers need to produce 'pushed output' to notice gaps in their knowledge and test hypotheses. When you rely on intermediaries for phone calls, you forfeit exactly the demanding output practice that drives productive skill development. Each avoided call prevents experiences that would recalibrate threat predictions: misunderstanding is survivable. Repair works. You can ask for repetition.

The research points to several strategies with varying levels of evidence support. Pre-task planning has the strongest evidence base for phone anxiety specifically. Task repetition reliably builds fluency. Graduated exposure follows established anxiety-treatment principles, though direct trials in L2 phone contexts are absent.

Pre-task planning is the best-evidenced intervention applicable to L2 phone anxiety. Even one minute of planning significantly improves accuracy. For phone calls specifically, the practical implication is clear: before calling, pre-formulate key phrases, anticipate vocabulary needs, and reduce the cognitive load of real-time production. Mak (2011) found that 'speaking without preparation' was the most anxiety-provoking factor — planning directly addresses this.

You do not need to eliminate anxiety. You need to function despite it. Here is a practical routine that operationalizes the research on pre-task planning, script preparation, and graduated exposure. Use it before high-stakes calls.

The goal is not a perfect call. The goal is a functional call. Broken German that gets the appointment scheduled moves the story forward. The research on willingness to communicate (MacIntyre et al., 1998) shows that some learners with high proficiency refuse to speak, while others with minimal knowledge communicate whenever possible. Be the second type.

The most striking finding of this review is how little direct research exists at the intersection of L2 anxiety and telephone communication specifically. The field needs:

• A validated telephone-specific L2 anxiety scale. Neither the PRCA-24 nor the FLCAS contains phone-specific items. No instrument captures medium-specific concerns: unpredictability of caller identity, inability to prepare environmental context, compensatory hypervigilance to paralinguistic cues.
• Experimental studies comparing L2 performance across communication modalities. No peer-reviewed study directly compares phone versus face-to-face L2 comprehension or production with matched tasks. All evidence uses video-versus-audio-only as proxy.
• Intervention trials targeting L2 phone-call anxiety specifically. Graduated exposure and systematic desensitization are well-established for general anxiety, but no treatment study applies them specifically to L2 phone contexts.
• German and French-specific research. Despite the prominence of these languages, direct empirical comparison of phone call anxiety in German or French L2 learners appears absent from the published literature.

The theoretical case for why L2 phone calls provoke disproportionate anxiety is robust. Visual cue removal increases processing demands. Low perceived control amplifies anxiety. Low self-efficacy intensifies the fear. Avoidance reinforces the cycle through negative reinforcement while depriving you of the output practice essential for development.

Yet the mechanisms operate at multiple levels simultaneously — and that means multiple points of intervention. Pre-task planning addresses unpredictability. Script preparation addresses the lack of target-language phone scripts. Task repetition builds procedural knowledge. Graduated exposure reduces conditioned fear through controlled experience.

You will not eliminate the anxiety. But you can make the call anyway. The research is clear: the practice you gain from functioning despite anxiety is what eventually reduces the anxiety itself. Not the other way around.

This article synthesizes findings from eye-tracking research (visual cue dependence), meta-analyses (anxiety-control relationships, FLA-self-efficacy), cognitive load theory, and task-based language teaching (planning, repetition). Links go to publisher pages (usually DOI).

1. Birulés J, Bosch L, Pons F, Lewkowicz DJ (2020) Attention to the mouth across auditory and visual contexts in monolingual and bilingual infants and adults. Language, Cognition and Neuroscience.
   Eye-tracking: L2 speakers attend significantly more to the talker's mouth when processing L2 vs. L1.
2. Grüter T, Pons F, Parlato-Oliveira E, Hiroshima K, Lee K, Fourlinnie I (2023) Visual attention to the mouth during L2 listening. Studies in Second Language Acquisition.
   Lower-proficiency L2 listeners fixate even more on the mouth — a gradient effect.
3. Sueyoshi A, Hardison DM (2005) The role of gestures and facial cues in second language listening comprehension. Language Learning.
   N=42 ESL learners: significantly better comprehension with visual cues at both proficiency levels.
4. Kwon SK, Yu G (2024) The effect of viewing visual cues in a listening comprehension test on L2 learners' test-taking process and performance: An eye-tracking study. Language Testing.
   N=57 Korean EFL learners with eye-tracking: examines how L2 listeners use visual cues during video-based listening tests.
5. Batty AO (2015) A comparison of video- and audio-mediated listening tests with many-facet Rasch modeling. Language Testing.
   N=200+: Small, non-significant differences between video and audio-only using many-facet Rasch modeling.
6. Kamiya N (2025) The limited effects of visual and audio modalities on second language listening comprehension. Language Teaching Research.
   N=52: Limited effects of watching gestures and lip movement on L2 listening comprehension.
7. Gallagher MW, Bentley KH, Barlow DH (2014) Perceived control and vulnerability to anxiety disorders. Cognitive Therapy and Research.
   Meta-analysis of 51 studies (N=11,218): large negative association between perceived control and anxiety.
8. Zhou J, Chiu MM, Dong Z, Zhou B (2023) The relationship between foreign language anxiety and self-efficacy: A meta-analysis. Current Psychology.
   Meta-analysis of 37 studies (N=26,589): r = −.70 between FLA and self-efficacy.
9. Kim JS, Oh HJ (2023) Telephone anxiety and digital communication preferences. Communication Research Reports.
   N=520: L2 status amplifies the relationship between digital technology use and telephone anxiety.
10. Vervliet B, Indekeu E (2015) Low-cost avoidance behaviors are resistant to fear extinction. Frontiers in Behavioral Neuroscience.
    Phone call avoidance is low-cost avoidance — resistant to extinction even after fear decreases.
11. Van Uijen SL, van der Linden D, Schmeets PMJ, Cremers HR, Emmerik REA (2017) Avoidance behavior predicts general anxiety 18 months later. PLOS ONE.
    N=221: Avoidance at baseline predicted higher anxiety at 18-month follow-up (b = 1.377, p < .001).
12. Wu X, Ellis R (2023) The effects of pre-task planning on L2 oral production. Language Learning Journal.
    N=43: Very large effect on fluency — r = .807, ηp² = .763.
13. Lambert C, Kormos J, Minn D (2017) Task repetition and L2 speech production. Studies in Second Language Acquisition.
    N=32: Largest fluency gains across first three performances, continued through fifth.
14. Abdi Tabari M, Zhuang J, Farahanynia M (2025) Task repetition effects on L2 performance: A meta-analysis. System.
    Meta-analysis: medium effect on syntactic complexity (d = 0.67), positive effects on accuracy and fluency.
15. Chen Y, Chew SY (2021) Speaking performance and anxiety levels in face-to-face and synchronous voice chat. Computer Assisted Language Learning.
    N=40 Chinese EFL learners: lower anxiety in audio-only — but context (classroom safety) matters critically.
16. Lindberg E, McDonough K, Trofimovich P (2022) Physiological anxiety in L2 conversation. Studies in Second Language Acquisition.
    N=60 with GSR monitoring: physiological arousal correlates with negative self-perceptions of fluency.
17. Bárkányi Z, Brash B (2025) Foreign language speaking anxiety, mental health, and online learning. Language Teaching (Cambridge Core).
    Systematic review: vocabulary retrieval is the primary online anxiety trigger; ~18% of learners cite it.
18. Divi C, Koss RG, Schmaltz SP, Loeb JM (2007) Language proficiency and adverse events in US hospitals. International Journal for Quality in Health Care.
    N=1,083 adverse event reports: ~50% of LEP adverse events resulted in physical harm vs. ~30% for English speakers.
19. MacIntyre PD, Dörnyei Z, Clément R, Noels KA (1998) Conceptualizing willingness to communicate in a L2. The Modern Language Journal.
    Willingness to communicate: some learners with high proficiency refuse to speak; others with minimal knowledge communicate whenever possible.
20. Swain M (1995) Three functions of output in second language learning. In: Cook G, Seidlhofer B (eds) Principle and Practice in Applied Linguistics.
    Output hypothesis: speakers need 'pushed output' to notice gaps, test hypotheses, and engage in metalinguistic reflection.
21. Teimouri Y, Goetze J, Plonsky L (2019) Second language anxiety and achievement: A meta-analysis. Studies in Second Language Acquisition.
    Meta-analysis: k=97, N=19,933, r = −.36 between anxiety and achievement.
22. Sánchez L, Choi Y, Oh S, et al. (2023) Does modality matter? A meta-analysis of video-based L2 listening. System.
    Effects of video on L2 listening are contingent — moderated by task type, learner level, visual information type.
23. Drijvers L, Van Der Plas M, Özyürek A, Jensen O (2019) Native and non-native listeners show differential neural responses to multimodal speech. NeuroImage.
    Non-native listeners show greater reliance on gestural cues with distinct oscillatory dynamics.
24. Mak B (2011) An exploration of speaking-in-class anxiety with Chinese ESL learners. System.
    N=313: 'Speaking without preparation' was the most anxiety-provoking factor.
25. Varonis EM, Gass SM (1985) Non-native/non-native conversations: A model for negotiation of meaning. Applied Linguistics.
    NNS-NNS pairs engage in significantly more repair than NS-NS pairs — repair is more necessary but harder by phone.