A speech planning network for interactive language use.

During conversation, people take turns speaking by rapidly responding to their partners while simultaneously avoiding interruption1,2. Such interactions display a remarkable degree of coordination, as gaps between turns are typically about 200 milliseconds3-approximately the duration of an eyeblink4. These latencies are considerably shorter than those observed in simple word-production tasks, which indicates that speakers often plan their responses while listening to their partners2. Although a distributed network of brain regions has been implicated in speech planning5-9, the neural dynamics underlying the specific preparatory processes that enable rapid turn-taking are poorly understood. Here we use intracranial electrocorticography to precisely measure neural activity as participants perform interactive tasks, and we observe a functionally and anatomically distinct class of planning-related cortical dynamics. We localize these responses to a frontotemporal circuit centred on the language-critical caudal inferior frontal cortex10 (Broca's region) and the caudal middle frontal gyrus-a region not normally implicated in speech planning11-13. Using a series of motor tasks, we then show that this planning network is more active when preparing speech as opposed to non-linguistic actions. Finally, we delineate planning-related circuitry during natural conversation that is nearly identical to the network mapped with our interactive tasks, and we find this circuit to be most active before participant speech during unconstrained turn-taking. Therefore, we have identified a speech planning network that is central to natural language generation during social interaction.

The neural basis of language development: Changes in lateralization over age.

We have long known that language is lateralized to the left hemisphere (LH) in most neurologically healthy adults. In contrast, findings on lateralization of function during development are more complex. As in adults, anatomical, electrophysiological, and neuroimaging studies in infants and children indicate LH lateralization for language. However, in very young children, lesions to either hemisphere are equally likely to result in language deficits, suggesting that language is distributed symmetrically early in life. We address this apparent contradiction by examining patterns of functional MRI (fMRI) language activation in children (ages 4 through 13) and adults (ages 18 through 29). In contrast to previous studies, we focus not on lateralization per se but rather on patterns of left-hemisphere (LH) and right-hemisphere (RH) activation across individual participants over age. Our analyses show significant activation not only in the LH language network but also in their RH homologs in all of the youngest children (ages 4 through 6). The proportion of participants showing significant RH activation decreases over age, with over 60% of adults lacking any significant RH activation. A whole-brain correlation analysis revealed an age-related decrease in language activation only in the RH homolog of Broca's area. This correlation was independent of task difficulty. We conclude that, while language is left-lateralized throughout life, the RH contribution to language processing is also strong early in life and decreases through childhood. Importantly, this early RH language activation may represent a developmental mechanism for recovery following early LH injury.

Broca's area involvement in abstract and concrete word acquisition: tDCS evidence.

Broca's area in the left hemisphere of the human neocortex has been suggested as a major hub for acquisition, storage, and access of linguistic information, abstract words in particular. Direct causal evidence for the latter, however, is still scarce; filling this gap was the goal of the present study. Using transcranial direct current stimulation (tDCS) of Broca's region, we aimed to delineate the involvement of this area in abstract and concrete word acquisition. The experiment used a between-subject design and involved 15 min of anodal or cathodal tDCS over Broca's area, or a sham/placebo control condition. The stimulation procedure was followed by a contextual learning session, in which participants were exposed to new concrete and abstract words embedded into short five-sentence texts. Finally, a set of behavioural assessment tasks was run to assess the learning outcomes immediately after the training (Day 1) and with a 24-hour delay (Day 2). The results showed that participants recognised novel abstract words more accurately after both anodal and cathodal tDCS in comparison with the sham condition on Day 1, which was also accompanied by longer recognition times (presumably due to deeper lexico-semantic processing), supporting the role of Broca's region in acquisition of abstract semantics. They were also more successful when recalling concrete words after cathodal tDCS, which indicates a degree of Broca's area involvement in forming memory circuits for concrete words as well. A decrease in the accuracy of recall of word forms and their meanings, as well as in recognition, was observed for all stimulation groups and both types of semantics on Day 2. The results suggest that both anodal and cathodal tDCS of Broca's area improves immediate contextual learning of novel vocabulary, predominantly affecting abstract semantics.

The Dynamics of Language Network Interactions in Lexical Selection: An Intracranial EEG Study.

Speaking in sentences requires selection from contextually determined lexical representations. Although posterior temporal cortex (PTC) and Broca's areas play important roles in storage and selection, respectively, of lexical representations, there has been no direct evidence for physiological interactions between these areas on time scales typical of lexical selection. Using intracranial recordings of cortical population activity indexed by high-gamma power (70-150 Hz) modulations, we studied the causal dynamics of cortical language networks while epilepsy surgery patients performed a sentence completion task in which the number of potential lexical responses was systematically varied. Prior to completion of sentences with more response possibilities, Broca's area was not only more active, but also exhibited more local network interactions with and greater top-down influences on PTC, consistent with activation of, and competition between, more lexical representations. These findings provide the most direct experimental support yet for network dynamics playing a role in lexical selection among competing alternatives during speech production.

Species and individual differences and connectional asymmetry of Broca's area in humans and macaques.

To reveal the connectional specialization of the Broca's area (or its homologue), voxel-wise inter-species and individual differences, and inter-hemispheric asymmetry were respectively inspected in humans and macaques at both whole-brain connectivity and single tract levels. It was discovered that the developed connectivity blueprint approach is able to localize connectionally comparable voxels between the two species in Broca's area, whereas the quantitative differences between blueprints of locationally or connectionally corresponding voxels enable us to generate inter-hemispheric, inter-subject, and inter-species connectional variabilities, respectively. More importantly, the inter-species and inter-subject variabilities exhibited positive correlation in both two primates, and relatively higher variabilities were detected in the anatomically defined pars triangularis. By contrast, negative relationship was identified between the inter-species variability and hemispheric asymmetry in human brain. In particular, relatively higher asymmetry was revealed in the anatomically defined pars opercularis. Therefore, our novel findings demonstrated that pars triangularis, as compared to pars opercularis, might be a more active area during primate evolution, in which the brain connectivity and possible functions of pars triangularis show relatively higher degree in species specialization, yet lower in hemispheric specialization. Meanwhile, brain connectivity and possible functions of pars opercularis manifested an opposite pattern. At the tract level, functional roles related to the ventral stream in speech comprehension were relatively conservative and bilaterally organized, while those related to the dorsal stream in speech production show relatively higher species and hemispheric specializations.

Decoding verbal working memory representations of Chinese characters from Broca's area.

Representations of sensory working memory can be found across the entire neocortex. But how are verbal working memory (VWM) contents retained in the human brain? Here we used fMRI and multi-voxel pattern analyses to study Chinese native speakers (15 males, 13 females) memorizing Chinese characters. Chinese characters are uniquely suitable to study VWM because verbal encoding is encouraged by their complex visual appearance and monosyllabic pronunciation. We found that activity patterns in Broca's area and left premotor cortex carried information about the memorized characters. These language-related areas carried (1) significantly more information about cued characters than those not cued for memorization, (2) significantly more information on the left than the right hemisphere and (3) significantly more information about Chinese symbols than complex visual patterns which are hard to verbalize. In contrast, early visual cortex carries a comparable amount of information about cued and uncued stimuli and is thus unlikely to be involved in memory retention. This study provides evidence for verbal working memory maintenance in a distributed network of language-related brain regions, consistent with distributed accounts of WM. The results also suggest that Broca's area and left premotor cortex form the articulatory network which serves articulatory rehearsal in the retention of verbal working memory contents.

Neurophysiological and functional neuroanatomical coding of statistical and deterministic rule information during sequence learning.

Humans are capable of acquiring multiple types of information presented in the same information stream. It has been suggested that at least two parallel learning processes are important during learning of sequential patterns-statistical learning and rule-based learning. Yet, the neurophysiological underpinnings of these parallel learning processes are not fully understood. To differentiate between the simultaneous mechanisms at the single trial level, we apply a temporal EEG signal decomposition approach together with sLORETA source localization method to delineate whether distinct statistical and rule-based learning codes can be distinguished in EEG data and can be related to distinct functional neuroanatomical structures. We demonstrate that concomitant but distinct aspects of information coded in the N2 time window play a role in these mechanisms: mismatch detection and response control underlie statistical learning and rule-based learning, respectively, albeit with different levels of time-sensitivity. Moreover, the effects of the two learning mechanisms in the different temporally decomposed clusters of neural activity also differed from each other in neural sources. Importantly, the right inferior frontal cortex (BA44) was specifically implicated in visuomotor statistical learning, confirming its role in the acquisition of transitional probabilities. In contrast, visuomotor rule-based learning was associated with the prefrontal gyrus (BA6). The results show how simultaneous learning mechanisms operate at the neurophysiological level and are orchestrated by distinct prefrontal cortical areas. The current findings deepen our understanding on the mechanisms of how humans are capable of learning multiple types of information from the same stimulus stream in a parallel fashion.

Functional brain plasticity during L1 training on complex sentences: Changes in gamma-band oscillatory activity.

The adult human brain remains plastic even after puberty. However, whether first language (L1) training in adults can alter the language network is yet largely unknown. Thus, we conducted a longitudinal training experiment on syntactically complex German sentence comprehension. Sentence complexity was varied by the depth of the center embedded relative clauses (i.e., single or double embedded). Comprehension was tested after each sentence with a question on the thematic role assignment. Thirty adult, native German speakers were recruited for 4 days of training. Magnetoencephalography (MEG) data were recorded and subjected to spectral power analysis covering the classical frequency bands (i.e., theta, alpha, beta, low gamma, and gamma). Normalized spectral power, time-locked to the final closure of the relative clause, was subjected to a two-factor analysis ("sentence complexity" and "training days"). Results showed that for the more complex sentences, the interaction of sentence complexity and training days was observed in Brodmann area 44 (BA 44) as a decrease of gamma power with training. Moreover, in the gamma band (55-95 Hz) functional connectivity between BA 44 and other brain regions such as the inferior frontal sulcus and the inferior parietal cortex were correlated with behavioral performance increase due to training. These results show that even for native speakers, complex L1 sentence training improves language performance and alters neural activities of the left hemispheric language network. Training strengthens the use of the dorsal processing stream with working-memory-related brain regions for syntactically complex sentences, thereby demonstrating the brain's functional plasticity for L1 training.

Hierarchical syntactic processing is beyond mere associating: Functional magnetic resonance imaging evidence from a novel artificial grammar.

Grammar is central to any natural language. In the past decades, the artificial grammar of the An Bn type in which a pair of associated elements can be nested in the other pair was considered as a desirable model to mimic human language syntax without semantic interference. However, such a grammar relies on mere associating mechanisms, thus insufficient to reflect the hierarchical nature of human syntax. Here, we test how the brain imposes syntactic hierarchies according to the category relations on linearized sequences by designing a novel artificial "Hierarchical syntactic structure-building Grammar" (HG), and compare this to the An Bn grammar as a "Nested associating Grammar" (NG) based on multilevel associations. Thirty-six healthy German native speakers were randomly assigned to one of the two grammars. Both groups performed a grammaticality judgment task on auditorily presented word sequences generated by the corresponding grammar in the scanner after a successful explicit behavioral learning session. Compared to the NG group, we found that the HG group showed a (a) significantly higher involvement of Brodmann area (BA) 44 in Broca's area and the posterior superior temporal gyrus (pSTG); and (b) qualitatively distinct connectivity between the two regions. Thus, the present study demonstrates that the build-up process of syntactic hierarchies on the basis of category relations critically relies on a distinctive left-hemispheric syntactic network involving BA 44 and pSTG. This indicates that our novel artificial grammar can constitute a suitable experimental tool to investigate syntax-specific processes in the human brain.

Functional neuroanatomy of language without speech: An ALE meta-analysis of sign language.

Sign language (SL) conveys linguistic information using gestures instead of sounds. Here, we apply a meta-analytic estimation approach to neuroimaging studies (N = 23; subjects = 316) and ask whether SL comprehension in deaf signers relies on the same primarily left-hemispheric cortical network implicated in spoken and written language (SWL) comprehension in hearing speakers. We show that: (a) SL recruits bilateral fronto-temporo-occipital regions with strong left-lateralization in the posterior inferior frontal gyrus known as Broca's area, mirroring functional asymmetries observed for SWL. (b) Within this SL network, Broca's area constitutes a hub which attributes abstract linguistic information to gestures. (c) SL-specific voxels in Broca's area are also crucially involved in SWL, as confirmed by meta-analytic connectivity modeling using an independent large-scale neuroimaging database. This strongly suggests that the human brain evolved a lateralized language network with a supramodal hub in Broca's area which computes linguistic information independent of speech.

The topographical organization of motor processing: An ALE meta-analysis on six action domains and the relevance of Broca's region.

Action is a cover term used to refer to a large set of motor processes differing in domain specificities (e.g. execution or observation). Here we review neuroimaging evidence on action processing (N = 416; Subjects = 5912) using quantitative Activation Likelihood Estimation (ALE) and Meta-Analytic Connectivity Modeling (MACM) approaches to delineate the functional specificities of six domains: (1) Action Execution, (2) Action Imitation, (3) Motor Imagery, (4) Action Observation, (5) Motor Learning, (6) Motor Preparation. Our results show distinct functional patterns for the different domains with convergence in posterior BA44 (pBA44) for execution, imitation and imagery processing. The functional connectivity network seeding in the motor-based localized cluster of pBA44 differs from the connectivity network seeding in the (language-related) anterior BA44. The two networks implement distinct cognitive functions. We propose that the motor-related network encompassing pBA44 is recruited when processing movements requiring a mental representation of the action itself.

Individual Differences in Reading Skill Are Related to Trial-by-Trial Neural Activation Variability in the Reading Network.

Recent work has suggested that variability in levels of neural activation may be related to behavioral and cognitive performance across a number of domains and may offer information that is not captured by more traditional measures that use the average level of brain activation. We examined the relationship between reading skill in school-aged children and neural activation variability during a functional MRI reading task after taking into account average levels of activity. The reading task involved matching printed and spoken words to pictures of items. Single trial activation estimates were used to calculate the mean and standard deviation of children's responses to print and speech stimuli; multiple regression analyses evaluated the relationship between reading skill and trial-by-trial activation variability. The reliability of observed findings from the discovery sample (n = 44; ages 8-11; 18 female) was then confirmed in an independent sample of children (n = 32; ages 8-11; 14 female). Across the two samples, reading skill was positively related to trial-by-trial variability in the activation response to print in the left inferior frontal gyrus pars triangularis. This relationship held even when accounting for mean levels of activation. This finding suggests that intrasubject variability in trial-by-trial fMRI activation responses to printed words accounts for individual differences in human reading ability that are not fully captured by traditional mean levels of brain activity. Furthermore, this positive relationship between trial-by-trial activation variability and reading skill may provide evidence that neural variability plays a beneficial role during early reading development.SIGNIFICANCE STATEMENT Recent work has suggested that neural activation variability, or moment-to-moment changes in the engagement of brain regions, is related to individual differences in behavioral and cognitive performance across multiple domains. However, differences in neural activation variability have not yet been evaluated in relation to reading skill. In the current study, we analyzed data from two independent groups of children who performed an fMRI task involving reading and listening to words. Across both samples, reading skill was positively related to trial-by-trial variability in activation to print stimuli in the left inferior frontal gyrus pars triangularis, even when accounting for the more conventional measure of mean levels of brain activity. This finding suggests that neural variability could be beneficial in developing readers.

Universal neural basis of structure building evidenced by network modulations emerging from Broca's area: The case of Chinese.

The basic steps in building up language involve binding words of different categories into a hierarchical structure. To what extent these steps are universal or differ across languages is an open issue. Here we examine the neural dynamics of phrase structure building in Chinese-a language that in contrast to other languages heavily depends on contextual semantic information. We used functional magnetic resonance imaging and dynamic causal modeling to identify the relevant brain regions and their dynamic relations. Language stimuli consisted of syntax-driving determiners, semantics-embedded classifiers, and nonverbal symbols making up for two-component sequences manipulated by the factors structure (phrase/list) and number of words (2-word/1-word). Processing phrases compared with word lists elicited greater activation in the anterior part of Broca's area, Brodmann area (BA) 45, and the left posterior superior/middle temporal gyri (pSTG/pMTG), while processing two words against one word led to stronger involvement of the left BA 45, BA 44, and insula. Differential network modulations emerging from subparts of Broca's area revealed that phrasal construction in particular highly modulated the direct connection from BA 44 to left pMTG, suggesting BA 44's primary role in phrase structure building. Conversely, the involvement of BA 45 rather appears sensitive to the reliance on lexico-semantic information in Chinese. Against the background of previous findings from other languages, the present results indicate that phrase structure building has a universal neural basis within the left fronto-temporal network. Most importantly, they provide the first evidence demonstrating that the structure-building network may be modulated by language-specific characteristics.

Neural substrates of word category information as the basis of syntactic processing.

The ability to use word category information (WCI) for syntactic structure building has been hypothesized to be the essence of human language faculty. The neural substrate of the ability of using the WCI for the complex syntactic hierarchical structure processing, however, is yet unknown. Therefore, we directly conducted an fMRI experiment by using a pseudo-Chinese artificial language with syntactic structures containing a center-embedded relative clause. Thirty non-Chinese native (Korean) speakers were randomly divided into two groups: one acquired WCI and WCI-based syntactic rules (the WCI group) before the scanning session, and the other did not (the non-WCI group). Both groups were required to judge the grammaticality of the testing sentences, with critical long-distance dependencies between two elements (the main verb and the relativizer). Behaviorally, the WCI group's accuracy was significantly higher and its reaction time was shorter. The scanning results showed that the left superior temporal gyrus (STG) and Broca's area were more strongly activated for the WCI group, and the dynamic causal modeling analyses revealed a distinct effective connectivity pattern for this group. Therefore, the present research, for the first time, reveals that the activation of and the functional connectivity between Broca's area and the left STG play a critical role in the ability of the rule-based use of the WCI which is crucial for complex hierarchical structure building, and might be substantially corresponding to the "labeling competence" within the linguistic framework.

Task load modulates tDCS effects on language performance.

Transcranial direct current stimulation (tDCS) effects in cognition are inconsistent across studies. This study aimed to discuss why typical models might be insufficient to explain these effects, and to investigate a brain state factor, task load, with behavioral experiments on phonological processing. The motor theory of speech perception states that motor codes for articulation take part in speech perception, a view sharpened by neuroimaging findings, which show that the motor role in phonological processing is weighted by the nature of the tasks. Three groups of 20 participants, each under a different tDCS condition (anodal, cathodal, or sham), performed a categorical perception (CP), a lexical decision (LD), and a word naming (WN) task while stimulated on the pars opercularis of the left inferior frontal gyrus, a language area typically involved with the motor role. These tasks were assumed to be subserved by a network of nodes which included the target, believed to be increasingly relevant for performance from speech perception to speech production. A-tDCS facilitation and C-tDCS downregulation should directly increase with the relevance of the target for the task. Downregulation of a low relevance node could result in facilitation by compensation from other nodes. Overall, our brain stimulation findings support the neuroimaging literature in that motor participation in phonological processing depends on task nature and show that tDCS effects are modulated by task load relative to the target. Outcomes such as the improved performance following cathodal tDCS in CP and WN suggest that compensatory mechanisms may take place when the tasks involve more complex neuronal networks.

A Parisian spring: the debate on language localization at the Imperial Academy of Medicine, Paris, April 4-June 13, 1865.

The localization of articulate language (speech) to the posterior third of the third left frontal convolution-Broca's area-did not occur to Broca as he reported the case of his first aphasic patient in 1861. Initially Broca localized articulate language to both frontal lobes, a position that he maintained for 4 years after publishing his first case. In the interval, the Academy of Medicine in Paris had received a copy of a paper authored in 1836 by Marc Dax, in which Dax claimed that the ability to speak resides within the left hemisphere alone. The Academy of Medicine convened in the spring of 1865 to adjudicate the issue. All of the distinguished speakers argued against Dax's contention by citing the prevailing paradigm, that bilaterally symmetrical organs, such as the eyes and ears, and the hemispheres of the brain, must perform the same function. The lone dissenting voice was that of Jules Baillarger, the discoverer of the laminar organization of the cerebral cortex, whose argument in favor of what he called "Dax's law" was so lucid that it carried the day. During his address to the Academy, Baillarger not only supported left-hemisphere dominance for speech, but for the first time described two forms of aphasia, fluent and nonfluent, now referred to as Wernicke's and Broca's aphasias, respectively, as well as the ability of aphasics to speak during emotional outbursts, to which we now refer as Baillarger-Jackson aphasia. It was 9 days after Baillarger's address that Broca, for the first time, unequivocally localized speech to the left frontal lobe.This paper is based on the author's reading of Dax's and Broca's original texts and of the texts read before the Academy of Medicine meeting held at the National Library of France between April 4, 1865, and June 13, 1865. From these primary sources it is concluded that the Academy of Medicine's debate was the last serious challenge to left-hemisphere dominance for speech and to the localization of articulate language to the left frontal lobe-and that Jules Baillarger played a pivotal role in what was a defining moment in neurobiology.

Mapping language dominance through the lens of the Wada test.

The sodium amytal test, or Wada test, named after Juhn Wada, has remained a pillar of presurgical planning and is used to identify the laterality of the dominant language and memory areas in the brain. What is perhaps less well known is that the original intent of the test was to abort seizure activity from an affected hemisphere and also to protect that hemisphere from the effects of electroconvulsive treatment. Some 80 years after Paul Broca described the frontal operculum as an essential area of expressive language and well before the age of MRI, Wada used the test to determine language dominance. The test was later adopted to study hemispheric memory dominance but was met with less consistent success because of the vascular anatomy of the mesial temporal structures. With the advent of functional MRI, the use of the Wada test has narrowed to application in select patients. The concept of selectively inhibiting part of the brain to determine its function, however, remains crucial to understanding brain function. In this review, the authors discuss the rise and fall of the Wada test, an important historical example of the innovation of clinicians in neuroscience.

Electrophysiological modulation and cognitive-verbal enhancement by multi-session Broca's stimulation: a quantitative EEG transcranial direct current stimulation based investigation.

To evaluate transcranial direct current stimulation-induced changes in resting state quantitative EEG and cognitive-verbal performance of second language learners, 16 healthy individuals were randomly recruited to sham and real transcranial direct current stimulation groups receiving eight sessions of second language instruction accompanied by a 2 mA transcranial direct current stimulation over Broca's area with the cathode placed over the left arm. Quantitative EEG was recorded during the resting state after the stimulation session and second language instruction. Reduced theta activity at Fp1, F7, F3, and T5 caused by the stimulus current was reported. Multisession stimulation resulted in a significant increase in current density for beta power (25 Hz) in the language network. Cognitive-verbal pre-post stimulation performances suggest that anodal vs. sham transcranial direct current stimulation significantly improved the subjects test score on digit span, a cognitive-verbal ability. It is concluded that transcranial direct current stimulation of Broca's area increase cognitive-verbal performance by modulating brain electrical activity in language-related regions.

Fundamental or forgotten? Is Pierre Paul Broca still relevant in modern neuroscience?

The ability to speak is a unique human capacity, but where is it located in our brains? This question is closely connected to the pioneering work of Pierre Paul Broca in the 1860s. Based on post-mortem observations of aphasic patients' brains, Broca located language production in the 3rd convolution of the left frontal lobe and thus reinitiated the localizationist view of brain functions. However, contemporary neuroscience has partially rejected this view in favor of a network-based perspective. This leads to the question, whether Broca's findings are still relevant today. In this mini-review, we discuss current and historical implications of Broca's work by focusing on his original contribution and contrasting it with contemporary knowledge. Borrowing from Broca's famous quote, our review shows that humans indeed "speak with the left hemisphere"- but Broca's area is not the sole "seat of articulatory language".

Perturbation of left posterior prefrontal cortex modulates top-down processing in sentence comprehension.

Communication is an inferential process. In particular, language comprehension constantly requires top-down efforts, as often multiple interpretations are compatible with a given sentence. To assess top-down processing in the language domain, our experiment employed ambiguous sentences that allow for multiple interpretations (e.g., The client sued the murderer with the corrupt lawyer., where the corrupt lawyer could either belong to The client or the murderer). Interpretation thus depended on whether participants chunk the words of the sentence into short or long syntactic phrases. In principle, bottom-up acoustic information (i.e., the presence or absence of an intonational phrase boundary at the offset of the murderer) indicates one of the two possible interpretations. Yet, acoustic information often indicates interpretations that require words to be chunked into overly long phrases that would overburden working memory. Processing is biased against these demands, reflected in a top-down preference to chunk words into short rather than long phrases. It is often proposed, but also hotly debated, that the ability to chunk words into short phrases is subserved by the left inferior frontal gyrus (IFG). Here, we employed focal repetitive transcranial magnetic stimulation to perturb the left IFG, which resulted in a further decrease of the aptitude to tolerate long phrases, indicating the inability of the left IFG to assist the chunking of words into phrases. In contrast, the processing of auditory information was not affected. Our findings support a causal top-down role of the left inferior frontal gyrus in the chunking of words into phrases.