CoChat

Age effects in L2 grammar processing as revealed by ERPs and how (not) to study them.

Open access

The DOI matches its Crossref record; 65 of the 65 references it cites were found in a registry or an index.

Abstract

In this study we investigate the effect of age of acquisition (AoA) on grammatical processing in second language learners as measured by event-related brain potentials (ERPs). We compare a traditional analysis involving the calculation of averages across a certain time window of the ERP waveform, analyzed with categorical groups (early vs. late), with a generalized additive modeling analysis, which allows us to take into account the full range of variability in both AoA and time. Sixty-six Slavic advanced learners of German listened to German sentences with correct and incorrect use of non-finite verbs and grammatical gender agreement. We show that the ERP signal depends on the AoA of the learner, as well as on the regularity of the structure under investigation. For gender agreement, a gradual change in processing strategies can be shown that varies by AoA, with younger learners showing a P600 and older learners showing a posterior negativity. For verb agreement, all learners show a P600 effect, irrespective of AoA. Based on their behavioral responses in an offline grammaticality judgment task, we argue that the late learners resort to computationally less efficient processing strategies when confronted with (lexically determined) syntactic constructions different from the L1. In addition, this study highlights the insights the explicit focus on the time course of the ERP signal in our analysis framework can offer compared to the traditional analysis.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1371/journal.pone.0143328

Cites

65 of 65 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig 1 Various age curves. Schematic representations of L2 proficiency (vertical axis) declining over AoA (horizontal axis). Only Panels B and C support a potential ‘critical period’, because of the presence of a discontinuity in the AoA function. Figure taken from [36].
Fig 1 Various age curves. Schematic representations of L2 proficiency (vertical axis) declining over AoA (horizontal axis). Only Panels B and C support a potential ‘critical period’, because of the presence of a discontinuity in the AoA function. Figure taken from [36].CC BY · PubMed Central
Fig 2 Group differences in multiple aspects of the ERP signal. An example of how ERP signatures can differ between groups (in this case native speakers, beginning and intermediate L2 learners). We see variation in form, latency and amplitude of parts of the ERP wave across these groups. It is difficult to objectively capture these differences in a traditional analysis, which is often artificially limited to time windows. Figure taken from [23].
Fig 2 Group differences in multiple aspects of the ERP signal. An example of how ERP signatures can differ between groups (in this case native speakers, beginning and intermediate L2 learners). We see variation in form, latency and amplitude of parts of the ERP wave across these groups. It is difficult to objectively capture these differences in a traditional analysis, which is often artificially limited to time windows. Figure taken from [23].CC BY · PubMed Central
Fig 3 Example of a basis function: a cubic regression spline. A cubic regression spline is a non-linear curve constructed from sections of cubic polynomials joined together so that the curve is continuous. The cubic regression spline shown (dotted curve) is made up of seven sections of cubic polynomials (solid lines). The points at which they are joined (and the two end points) are known as the ‘knots’ of the spline. Figure taken from [43].
Fig 3 Example of a basis function: a cubic regression spline. A cubic regression spline is a non-linear curve constructed from sections of cubic polynomials joined together so that the curve is continuous. The cubic regression spline shown (dotted curve) is made up of seven sections of cubic polynomials (solid lines). The points at which they are joined (and the two end points) are known as the ‘knots’ of the spline. Figure taken from [43].CC BY · PubMed Central
Fig 4 Location of the electrodes and ROI. Approximate location of the EEG recording sites and the central-posterior region of interest used for analysis.
Fig 4 Location of the electrodes and ROI. Approximate location of the EEG recording sites and the central-posterior region of interest used for analysis.CC BY · PubMed Central
Fig 5 Accuracy on (behavioral) grammaticality judgments made during the EEG recording session. Performance in percentage of accurate responses is plotted separately for each level of group (learners, natives), structure (verb, gender) and grammaticality (grammatical, ungrammatical).
Fig 5 Accuracy on (behavioral) grammaticality judgments made during the EEG recording session. Performance in percentage of accurate responses is plotted separately for each level of group (learners, natives), structure (verb, gender) and grammaticality (grammatical, ungrammatical).CC BY · PubMed Central
Fig 6 Grand average ERP responses. Waveforms show the ERP signal averaged over the electrodes in the central-posterior region of interest (see Fig 4) for natives and learners. The green lines represent the average response to grammatical targets, the red lines show the response to violations. Underneath the waveform graphs, voltage maps are plotted, which show the scalp topography of the effect in the difference wave (i.e. mean amplitude in ungrammatical minus grammatical condition) in the 0–500, 500–1000 and 1000–1400 ms time window. The scalp topographies are visualized looking down on the head (i.e. the location of the nose is at the top), with positive differences in red, negative differences in blue, and green representing no difference. A traditional group split divides learners into early (7–16) and late (17–36) age of onset of acquisition groups. Both natives and learners show P600 effects for non-finite verb agreement. Only early learners show the same effect for gender agreement, whereas late learners do not show sensitivity for gender violations.
Fig 6 Grand average ERP responses. Waveforms show the ERP signal averaged over the electrodes in the central-posterior region of interest (see Fig 4) for natives and learners. The green lines represent the average response to grammatical targets, the red lines show the response to violations. Underneath the waveform graphs, voltage maps are plotted, which show the scalp topography of the effect in the difference wave (i.e. mean amplitude in ungrammatical minus grammatical condition) in the 0–500, 500–1000 and 1000–1400 ms time window. The scalp topographies are visualized looking down on the head (i.e. the location of the nose is at the top), with positive differences in red, negative differences in blue, and green representing no difference. A traditional group split divides learners into early (7–16) and late (17–36) age of onset of acquisition groups. Both natives and learners show P600 effects for non-finite verb agreement. Only early learners show the same effect for gender agreement, whereas late learners do not show sensitivity for gender violations.CC BY · PubMed Central
Fig 7 The effect of AoA on the average ERP signal. The figure shows the average magnitude of the difference wave per individual (x-axis), against that individual’s AoA (y-axis). This magnitude was calculated by taking the ERP signal per condition, averaged over the electrodes in the central-posterior region of interest (see Fig 4) using the mean amplitude in the 500–1200 ms time window, to calculate the difference between the individual’s response to ungrammatical versus grammatical targets. The plots reveal a weak correlation between AoA and the magnitude of the grammaticality effect in the gender condition (r(64) = -0.32, p = .009), but no effect in the verb condition (r(64) = 0.03, p = .825).
Fig 7 The effect of AoA on the average ERP signal. The figure shows the average magnitude of the difference wave per individual (x-axis), against that individual’s AoA (y-axis). This magnitude was calculated by taking the ERP signal per condition, averaged over the electrodes in the central-posterior region of interest (see Fig 4) using the mean amplitude in the 500–1200 ms time window, to calculate the difference between the individual’s response to ungrammatical versus grammatical targets. The plots reveal a weak correlation between AoA and the magnitude of the grammaticality effect in the gender condition (r(64) = -0.32, p = .009), but no effect in the verb condition (r(64) = 0.03, p = .825).CC BY · PubMed Central
Fig 8 Visualization of the GAM model for the effects of time and AoA on the difference wave, for the verb (Panel A) and gender (Panel B) condition. First, the individual smooths and surfaces for the additive effects of time (Panels A1 and B1), AoA (Panels A2 and B2) and their pure interaction (Panels A3 and B3) are plotted separately. These are followed by a plot that summarizes all these effects together (the ‘full effect’ plots, Panels A4 and B4). After taking out the main effect of time, the effect of AoA is not significant on its own, for either verb or gender, but only in interaction with time (see Table 4). These adjustments to the general time pattern match the complexity of the AoA effects on timing, slope and polarity of the ERP components seen in the ‘full effect’ plots: In the verb condition, we see a P600 effect across all AoAs, with only minor adjustments to amplitude, onset and duration of this component for earlier vs. later learners. For gender, there is a diminishing P600 effect for early AoAs and instead a negativity that peaks around 500 ms for later AoAs.
Fig 8 Visualization of the GAM model for the effects of time and AoA on the difference wave, for the verb (Panel A) and gender (Panel B) condition. First, the individual smooths and surfaces for the additive effects of time (Panels A1 and B1), AoA (Panels A2 and B2) and their pure interaction (Panels A3 and B3) are plotted separately. These are followed by a plot that summarizes all these effects together (the ‘full effect’ plots, Panels A4 and B4). After taking out the main effect of time, the effect of AoA is not significant on its own, for either verb or gender, but only in interaction with time (see Table 4). These adjustments to the general time pattern match the complexity of the AoA effects on timing, slope and polarity of the ERP components seen in the ‘full effect’ plots: In the verb condition, we see a P600 effect across all AoAs, with only minor adjustments to amplitude, onset and duration of this component for earlier vs. later learners. For gender, there is a diminishing P600 effect for early AoAs and instead a negativity that peaks around 500 ms for later AoAs.CC BY · PubMed Central
Fig 9 The effect over time on the ERP signal for learners with an AoA of 10, 20 and 30 for verb and gender violations. These plots present an alternative visualization of the two-dimensional ‘full effect’ surfaces of Fig 8: AoA values, which were represented on the y-axis in Fig 8 Panels A4 and B4, are now represented by three separate plots for particular AoA values, with microvoltage on the y-axis. The gradual transitions towards different AoAs illustrate the continuity of the effect.
Fig 9 The effect over time on the ERP signal for learners with an AoA of 10, 20 and 30 for verb and gender violations. These plots present an alternative visualization of the two-dimensional ‘full effect’ surfaces of Fig 8: AoA values, which were represented on the y-axis in Fig 8 Panels A4 and B4, are now represented by three separate plots for particular AoA values, with microvoltage on the y-axis. The gradual transitions towards different AoAs illustrate the continuity of the effect.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 65 references

Research works printed but not linked: 3

  1. SN Wood, 2015, Generalized additive models for large datasets, J R Stat Soc Ser C Appl Stat
  2. G Corbett, 1991, Cambridge textbooks in linguistics
  3. G Gratton, 1989, Generalization and evaluation of eye-movement correction procedures, J Psychophysiol, 3, 14

Cited by

See all 66

Metadata

Authors
Meulman, N., Wieling, M., Sprenger, S. A., Stowe, L. A., & Schmid, M. S.
Affiliations
  • University of Groningen (NL)
  • University of Essex (GB)
Corresponding author
Meulman, Nienke
Venue
PLOS ONE
Published
2015-12-18
Volume
10
Issue
12
Pages
e0143328
ISSN
1932-6203 (ISSN-L)
OpenAlex
W2218032177
Cited by
60
References
63
Citations by year
2026: 3 · 2025: 9 · 2024: 4 · 2023: 10 · 2022: 4 · 2021: 6 · 2020: 6 · 2019: 2 · 2018: 7 · 2017: 4 · 2016: 4
FWCI
3.00
Type
journal article
Language
en
Open access
Locations
  • PLoS ONE (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Open Access at Essex (University of Essex) (acceptedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Figshare (submittedVersion, cc-by)
  • University of Groningen research database (University of Groningen / Centre for Information Technology) (submittedVersion, cc-by)
Funders
  • Universität Hamburg
  • Technische Universität Berlin
Topics
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Neurobiology of Language and Bilingualism
  • Social Sciences › Arts and Humanities › Language and Linguistics › EFL/ESL Teaching and Learning
  • Social Sciences › Psychology › Developmental and Educational Psychology › Language Development and Disorders
Keywords
age of acquisition, event-related potentials, P600, gender agreement, non-finite verbs, verb agreement, generalized additive models, grammaticality judgment
MeSH terms
Adult, Age Factors, Brain, Evoked Potentials, Female, Humans, Male, Middle Aged, Models, Biological, Vocabulary, Multilingualism, Comprehension, Young Adult

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Age effects in L2 grammar processing as revealed by ERPs and how (not) to study them. https://app.cochat.ai/papers/10.1371/journal.pone.0143328
The paper
Meulman, N., Wieling, M., Sprenger, S. A., Stowe, L. A., & Schmid, M. S. (2015). Age effects in L2 grammar processing as revealed by ERPs and how (not) to study them. PLOS ONE. https://doi.org/10.1371/journal.pone.0143328

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.