Edited by: Elena Salillas, University of Padua, Italy
Reviewed by: Chiara Finocchiaro, University of Trento, Italy; Suhan Senova, Assistance Publique Hopitaux De Paris, France
This article was submitted to Cognitive Neuroscience, a section of the journal Frontiers in Human Neuroscience
†These authors have contributed equally to this work
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Navigated Transcranial Magnetic Stimulation (nTMS) is used to understand the cortical organization of language in preparation for the surgical removal of a brain tumor. Action naming with finite verbs can be employed for that purpose, providing additional information to object naming. However, little research has focused on the properties of the verbs that are used in action naming tasks, such as their status as transitive (taking an object; e.g.,
Navigated Transcranial Magnetic Stimulation (nTMS) is used to delimit (i.e., map) the cortical representation of language in preparation for the removal of a brain tumor in the context of an awake surgery (Tarapore et al.,
At the clinical level, the presence of errors during nTMS is used to guide the surgical procedure by issuing recommendations on the basis of points that elicit a positive or negative reaction (Picht et al.,
Object naming has been used in nTMS studies to examine the representation of language in the brain (Picht et al.,
The design and administration of action naming tasks is similar to object naming tasks. In action naming tasks, participants are shown black and white drawings of a character or animal carrying out an action, and participants are asked to name the event using an infinitive
In this study, we explored a new approach to study nTMS data. We analyzed the types of verbs in an action naming task with a specific focus on the number of arguments that verbs can take (i.e., transitivity), specifically the difference between transitive (e.g.,
Verbs differ in terms of their syntactic properties regarding the number and type of syntactic complements they take (e.g., Chomsky,
(1) a. Mary is winking.
b. Mary is reading a book.
c. *Mary is winking a book.
d. Mary is reading.
e. *Mary is fixing.
f. Mary is fixing a bicycle.
Different cognitive models account for the influence of transitivity in single word and sentence production. Lexicalist approaches argue that grammatical information is stored within the lemma (i.e., syntactic properties and meaning of a word; Roelofs et al.,
Predictions of theoretical frameworks regarding transitivity have been confirmed by experimental evidence from individuals with agrammatic aphasia. In spontaneous speech, people with post-stroke agrammatic aphasia and Alzheimer's disease tend to produce relatively more intransitive verbs and fewer transitive verbs than non-brain damaged speakers (Bastiaanse and Jonkers,
In sum, reports on people with agrammatic aphasia indicate more difficulties with transitive than intransitive verbs either in spontaneous speech or action naming tasks. Some of the aforementioned studies indicate the location of neurological damage, typically in perisylvian areas of the left hemisphere (i.e., Jonkers,
fMRI studies have examined the influence of argument structure during sentence and single word comprehension, as well as single word production. Evidence from sentence comprehension connects left posterior temporal and inferior frontal regions with the processing of verbs that assign an increased number of arguments and thematic roles (Ben-Shachar et al.,
In a series of studies with non-brain-damaged participants, Thompson and colleagues employed lexical decision tasks. The authors reported that the processing of transitive verbs generates higher activation in inferior parietal regions of the left and right hemisphere (i.e., angular and supramarginal gyrus) compared to the processing of intransitive verbs (Thompson et al.,
Most research with fMRI has examined argument structure processing during comprehension. An exception is the study by den Ouden et al. (
TMS over the left inferior parietal lobe has also been shown to facilitate thematic role assignment, adding causal evidence to the role of posterior regions in argument structure processing (Finocchiaro et al.,
Summary of areas that have been reported to elicit more activation during the processing of transitive over intransitive verbs (blue areas) and those for the opposite contrast (i.e., more activation for intransitive compared to transitive verbs; light blue areas). Numbers indicate the 46 stimulation points used for the current study according to CPS regions (Corina et al.,
The aim of the present study is to explore the cortical representation of transitive and intransitive verbs with nTMS and to investigate whether transitivity affects the number and localization of nTMS-induced errors. Hence, we ask the following questions:
Does nTMS induce more (or fewer) errors with transitive compared to intransitive verbs?
If so, in which cortical regions (i.e., hemisphere and lobes) can we localize the nTMS-induced error rates for transitive and intransitive verbs?
Do transitive and intransitive verbs elicit different types of nTMS-induced errors?
Based on previous literature, we assumed that if transitive verbs generate larger cortical activity as seen in aforementioned neuroimaging studies, nTMS would induce more errors with transitive compared to intransitive verbs in posterior regions. Concerning specific error categories, we hypothesized that if the locus of complexity for transitive verbs compared to intransitive verbs is indeed the lexico-semantic level, then we will observe more errors of the lexico-semantic category.
We analyzed previously reported data by focusing solely on the comparison between transitive and intransitive verbs during action naming under nTMS (Ohlerth et al., submitted).
Twenty neurologically healthy participants were tested. They ranged in age 20–53 (mean age: 24.75, SD = 7). They were 12 females, 1 left-handed, and 1 ambidextrous individual. The inclusion criteria were: (1) German as a native language, (2) age of at least 18 years, (3) no contra-indications for Magnetic Resonance Imaging (MRI) 3 Tesla and/or nTMS mapping (i.e., use of cardiac pacemakers or devices for deep brain stimulation), (4) no neurological or psychiatric disorders, and (5) no pregnancy. Handedness was measured using the Edinburgh Handedness Inventory (Oldfield,
Anatomical T1-weighted MRI images were acquired using a 3-Tesla magnetic resonance scanner (Achieva dStream; Philips Healthcare, Best, The Netherlands). 3D models of each participant's brain were constructed based on the acquired MRI images. These models were used for the guidance of coil placement during language mapping with stimulation (Nexstim eXimia NBS system version 4.3).
The German version of the Verb And Noun test for Peri OPerative testing (VAN-POP; Ohlerth et al.,
Example of an intransitive
A focal figure-of-eight coil was used. It produced biphasic pulses (length 23 mm) with maximal electric field strength of 172 V/m ± 2%. Prior to mapping, T-1 weighted MRI sequences were uploaded to the Nexstim eXimia NBS system version 4.3. According to the Cortical Parcellation System (CPS; Corina et al.,
Surface electrodes for the recording of Motor Evoked Potentials were placed over the abductor pollicis brevis and abductor digiti minimi muscles to establish the Resting Motor Threshold (rMT) according to the preoperative language mapping protocol by Krieg et al. (
Participants were presented with the set of images one-by-one while seated ~60 cm in front of a computer screen. Picture presentation time (PPT) was set at 1,000 ms and inter-picture interval (IPI) at 2,500 ms. Participants were instructed to name the pictures as fast and accurately as possible. They were asked to overtly produce the lead-in phrase and the verb, while refraining from producing the object of transitive verbs (i.e.,
For each participant, two rounds of baseline naming were completed with the absence of any stimulation and misnamed items were excluded from the nTMS experiment for that participant. Hence, each person named an individualized set of images for the nTMS testing. This procedure is commonly used in preoperative language mapping (Krieg et al.,
The instructions for mapping with stimulation as well as IPI and PPT remained the same as during baseline naming. The interval between stimulation onset and picture presentation was set at 0 ms, so that picture and stimulation onset were synchronized. As it is common in language mapping protocols, we used repetitive stimulation with the intensity set at 5Hz/5 pulses, with a typical duration of 1,000 ms (Krieg et al.,
A trained clinical linguist with expertise in nTMS language mapping performed all nTMS mappings and analyzed the
The category of
We conducted quantitative analysis of the errors as well as a qualitative analysis according to error types. Quantitatively, to check for differences between transitive and intransitive verbs, we conducted chi-square tests for each hemisphere and lobe. The induced errors were then divided into different error types and chi-square tests were conducted for each hemisphere and lobe, to see whether the quality of errors differed between transitive and intransitive verbs. Significance values were corrected for False Discovery Rate (i.e., FDR; Benjamini and Hochberg,
nTMS stimulation induced errors in all participants. The mean number of items that were misnamed and, hence, excluded after baseline testing was 11.00 (sd = 4.4). The number of nTMS stimulations across participants was 11,040. Of these, 918 stimulations induced errors (8.3%). Errors were elicited in both hemispheres, with 429 errors occurring in the left hemisphere (3.9%) and 489 in the right (4.4%). A paired
Out of the 7,507 stimulations with transitive items, 661 stimulations elicited errors (8.8%). Out of the 3,533 stimulations with intransitive items, 257 stimulations induced errors (7.3%). This difference was significant (χ2 = 7.18,
Analyzing the effect of transitivity per hemisphere, significantly more nTMS-positive points were identified with transitive items in the left hemisphere (transitives = 9.2%; intransitives = 6.9%; χ2 = 5.49,
Error percentages of nTMS induced errors according to verb type and hemisphere. The axis has been set to 20% for purposes of visualization. *
Regarding our lobe-wise analysis, a significantly higher number of nTMS-induced errors was elicited during the production of transitive verbs compared to intransitive verbs when stimulating the left parietal lobe (transitives = 9%; intransitives = 5.4%; χ2 = 6.28,
The most frequently induced error category was lexico-semantic errors that accounted for 45.8% of all errors (
Results of the qualitative analysis across hemispheres (percentages of errors).
Non-linguistic errors | Hesitation on the sentence | Transitive |
16.04 |
No response | Transitive |
2.12 |
|
Lexico-semantic errors | Hesitation on the target | Transitive |
33.89 |
Anomia | Transitive |
5.45 |
|
Semantic | Transitive |
6.81 |
|
Grammatical errors | Grammatical | Transitive |
4.84 |
Sound level | Phonological | Transitive |
0.30 |
Performance | Transitive |
30.56 |
In the left hemisphere, there was no difference between the error categories for the transitive nor the intransitive verbs. When examining the separate lobes of the left hemisphere, no significant differences were found in the frontal and temporal lobes, but in the left parietal lobe, we found significantly more nTMS-induced lexico-semantic errors with transitive than with intransitive verbs (4.6 vs. 2.2%). A summary of the results according to error category and lobe can be found in
Error types elicited in the left hemisphere per lobe.
Frontal | χ2 =0.31, |
χ2 = 0.00, |
W = 0.35, |
χ2 = 0.50, |
Temporal | W = 0.28, |
χ2 = 0.00, |
– | W = −1.2, |
Parietal | χ2 = 0.11, |
χ2 = 5.38, |
– | χ2 = 0.45, |
No significant differences were found between the error types for transitive and intransitive verbs in the right hemisphere (see
In the present study, nTMS induced a larger number of errors with transitive compared to intransitive verbs. Examining each hemisphere separately, the error rate with transitive verbs was higher than that of intransitive verbs in the left hemisphere and particularly in the left parietal lobe. Qualitatively, more lexico-semantic errors with transitive items compared to intransitive items were produced during stimulation in the left parietal lobe. These findings will be discussed within the above-presented theoretical context and evidence from behavioral studies in people with aphasia, as well as in relation to neuroimaging studies.
Theories of argument structure predict that the complexity of transitive verbs is associated with linguistic processes either at the lemma or at the sentence level (Bock and Levelt,
Regarding the exact linguistic features that render transitive verbs more complex, the present study cannot offer resolution. As previously stated, we opted for an action-naming-in-sentence-context task to elicit verbs inflected for number, person, and tense (e.g.,
Previous studies that implemented sentence context in action naming either presented sentences with varying number of arguments (e.g., Ben-Shachar et al.,
However, the locus of increased complexity of transitive verbs compared to intransitives is not necessarily exclusively due to lemma retrieval. As previous studies on post-stroke aphasia have demonstrated, observed difficulties with verb production are due to impairments affecting the level of
Our findings stress the role of the left parietal lobe regarding argument structure. Parietal areas reported in previous neuroimaging studies were argued to function as a repository of information regarding argument structure (Thompson et al.,
In our study, nTMS induced more lexico-semantic errors with transitive items in the left parietal lobe. These errors are associated with the level of access to the lexicon (Corina et al.,
Apart from argument structure information retrieval, the left parietal lobe has also been connected with grammar. Evidence from Basque and English shows that syntactic anomalies, such as case or verb agreement violations, activate the left and right inferior parietal lobes, whereas semantic anomalies do not (Kuperberg et al.,
Although previous literature reported increased activation with transitive compared to intransitive verbs in right posterior regions, we did not observe more nTMS-induced errors for transitive compared to intransitive verbs in the right hemisphere (Thompson et al.,
First, previous studies have reported right hemisphere activation for experimental designs and conditions that differ from those in the present study. den Ouden et al. (
Since Broca's findings on patient Tan (Broca,
Within the context of transitivity, unlike previous literature that employed fMRI, non-correlational techniques, such as TMS, potentially affect necessary processes for verb production (Genon et al.,
Our findings highlight the importance of administering linguistically motivated naming tasks within the context of preoperative nTMS language mapping. Specifically, concerning action naming tasks, we suggest that the variable “transitivity” should be controlled. Regardless of the reasons behind the increased lexico-semantic complexity of transitive verbs, it becomes apparent that transitivity affects the number and localization of nTMS-positive points. Careful consideration of these variables during the construction of action naming tasks is, hence, important for language mapping with nTMS.
In the present study, we reported that an increased amount of argument structure information particularly affects the left parietal lobe. This finding indicates that when clinicians map peritumoral regions in left parietal areas with nTMS, they may opt for an action naming task with transitive items or at least consider such items in the task at hand. The use of exclusively intransitive items may lead to low numbers of induced errors, that may lead to the conclusion that these areas are not involved in language. This is also emphasized by a recent review of TMS and DES language mapping, which reported that the highest number of false-negative TMS points is located in posterior language areas (Jeltema et al.,
It should be noted that the exact number of stimulations per lobe with transitive and intransitive items was not available to the researchers. This is because the video analysis software of the present study is blinded to the stimulation site. Once an error has been identified, the stimulation sequence is stored and marked on the anatomical space. However, this does not apply to stimulations that did not induce errors. Hence, to conduct lobe-wise analyses, we assumed that the distribution of the presented transitive and intransitive items was equivalent to the overall distribution of our items (i.e., 68% transitive items and 32% intransitive). Our assumption is justified by the fact that the action naming task was presented 8-9 times per participant and was randomized every time after the final item. Several questions arise from our findings: Evidence from pre- and intraoperative studies with nTMS and DES in people with brain tumors can be used to cross-validate our findings and show whether similar effects can be found in individuals with neuro-pathologies. Additionally, it seems important to examine word properties of verbs (and nouns) within mapping with nTMS and DES. Large sets of nTMS data can be analyzed to identify word properties affecting naming accuracy and proneness to nTMS disruption (Alyahya et al.,
Our study shows that the number as well as type of nTMS-induced errors can be differentially affected based on the items of an action naming task. In particular, it provides causal evidence for the complexity of transitive verbs (vs. intransitives verbs) and their cortical representation in the left parietal lobe.
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
The studies involving human participants were reviewed and approved by the Ethics Committee of the Technical University of Munich. The patients/participants provided their written informed consent to participate in this study.
EN, A-KO, SI, SK, RB, and AR: conceptualization, methodology, validation, investigation, and writing—review and editing. EN, A-KO, and AR: formal analysis. SI and SK: resources and funding acquisition. EN and A-KO: data curation. EN, A-KO, AR, and RB: writing—original draft preparation. EN: visualization. SI, SK, RB, and AR: supervision and project administration. All authors contributed to the article and approved the submitted version.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
We are grateful to the participants for their time and interest in our study.
The Supplementary Material for this article can be found online at:
1This also applies to pseudo-transitive verbs, which also allow two arguments and comprise the verb type used in the present study.