Download A Review on Brain Mechanisms for Language Acquisition and Comprehension and more Summaries Psychology in PDF only on Docsity!
A Review on Brain Mechanisms for Language Acquisition and
Comprehension
Kailash Nath Tripathi
1
, Anand Bihari
2
, Sudhakar Tripathi
3
, Ravi Bhushan Mishra
4
1 Merrut Institute of Technology, Merrut, 2 Vellore Institute of Technology, Vellore, 3 R. E. C. Ambedkar Nagar,
U. P. 4 IIT (BHU) Varanasi
1
Abstract:
This paper aims to bring the main viewpoint of language acquisition and language comprehension. In language
acquisition, we have reviewed the different types of language acquisitions like first, second, sign and skill
acquisition. The experimental techniques for neurolinguistic acquisition detection is also discussed. The findings
of experiments for acquisition detection is also discussed, it includes the region of brain activated after
acquisition. Findings shows that the different types of acquisition involve different regions of the brain. In
language comprehension, native language comprehension and bilingual’s comprehension has been considered.
Comprehension involve different brain regions for different sentence or word comprehension depending upon
their semantic and syntax. The different fMRI/EEG analysis techniques (statistical/ graph theoretical) are also
discoursed in our review.Tools for neurolinguistic computations (pre-processing/computations/analysis) are
alsodiscussed.
Keywords: Language Acquisition, Language Comprehension, Cognition, GLM, ICA, PCA, ERP, t-test, z-score.
1 Introduction
The past several years has yielded an enormous research work in neuroscience investigating language
acquisition, comprehension and production. Non-invasive, safe functional brain measurements have now been
proven feasible for use with infants or adult for neural data acquisition. The neural signature of effect of learning
at the phonetic level can be recognized at a amazingly high precision. Continuity in linguistic development,
brain responses to even phonetic level stimuli can be observed with theoretical and clinical impact.
2Language Acquisitions
Human brain the command centre controls heart rhythm, memory and language to all human activities. Broca’s
area a small region in inferior frontal gyrus(IFG) necessary for production and coordination of language is found
in left hemisphere in most of people. Wernicke’s area the counter part of Broca’s area in superior temporal
gyrus(STG) performs language comprehension both written and spoken. The area of Broca’s area is
usually described as composed of the cytoarchitecturally defined area of Brodmann BA44, the pars opercularis
and BA 45, and the pars triangularis.The cytoarchitecturally identified region BA 22 covers the latter two-thirds
of the lateral convexity of the STG and is part of the Wernicke region.[1]
Figure 1: Langauge area in human brain comprises Broca’s and Wernicke’s Area
The acquisition of languages is one of the most important human traits and certainly it is the brain that
undergoes the changes in development. Therefore the root of grammatical rules should be ascribed to an
implicit process in the human brain. Linguists find speaking, signing and understanding language to be the key
language skills, i.e. natural or inborn and biologically determined, while they find reading and writing to be
secondary. In truth, acquisition of a native or first language (L1) through these primary faculties during the first
years of life, whereas children learn their linguistic knowledge gradually. Speech in children progresses from
babbling at around the age of 6-8 months, to the single-word stage at 10 to 12 months, and then to the two-word
stage at around 2 years. There is a profound difference between linguistic factors between L1 and L2. An
L2(Second Language) can be learned at any moment in life, although the L2 capacity is rarely comparable to
that of L1 if it is acquired after the predicted 'sensitive period’ from early childhood to puberty (È12 years of
age). Numerous studies of functional magnetic resonance imaging(fMRI) and positron emission tomography
(PET) have shown that auditory phonological processing is correlated with activation in the posterior superior
temporal gyrus (STG) [Brodmann's region (BA) 22], while lexico-semantic processing is typically associated
with activation in the left extra-Sylvian temporoparietal regions, including the angular ones [2].
In [3], Eric Lenneberg (1967) proposed that the acquisition of human language was an example of biologically
limited learning, he stated that a child would have biological heritable component to learn language. He
concluded that the process of acquiring langauge is profoundly ingrained and, species-specific, human
biological property. Any language usually acquired during a crucial time beginning early in life and ending in
puberty. He indicated that language could only be learned with difficulty or through a different learning method
beyond this time.
A critical period is a time of maturation during which some of the key stimuli would have their peak impact on
development or learning, resulting in normal actions adapted to the specific environment to which the organism
was exposed. If the organism is not subjected to this phenomenon until after this period of time, the same
phenomenon may have either a diminished effect, or may have no effect at all in extreme cases. Studies show a
close association between age of language use and the ultimate degree of competency (PL) attained. However,
exposure age does not affect all aspects of language leaning equally. Therefore, the crucial effects of the critical
period seem to focus on phonology, morphology, syntax and not meaning processing [4].
The existence of critical, or at least a sensitive period for language acquisition in human being is explained by
an evolutionary model suggested by J. R. Hurford in [5].
The acquisition of first language is one of the unexplained mysteries which surround us in our daily lives. A
child learns language spontaneously, almost miraculously, as its learning of language progresses rapidly with an
obvious pace and accuracy. Most children quickly learn language, giving the illusion that the process of
acquiring first language is easy and straightforward. This is not the case, however, as children go through many
stages of first language acquisition .The stages of language learning in children usually consist of: cooing,
babbling, holophrastic stage, telegraphic speech and normal speech. The age of cooing is up-to 9 months till
then children use phonemes from every language. At 9 month they start babbling in which they selectively use
phonemes from their native language. At the age of 12 month they start using single words. When they are in
holophrastic stage at around 18-24 months, they can combine words in two words stages.By the age of around
30 months they develop to the telegraphic stage where they can utter a clear phrase structure. As the children
develop physically, so does their language skills as they internalize more complex systems by widening their
vocabulary and their immediate surroundings. At the age of 5 years children reached up to normal developed
speech.
There are three famous theories for first language acquisition: the behaviourist theory, the innatist theory and
interactionists theory. Behaviourist theory[7] equated learning to a language all behaviour are acquired through
interaction with environment and interactions are imitation, reinforcement, practice and habit formation.
Children learn their first language by stimuli and children's responses are influenced by reinforcement.
The Innatist theory[8] believed that children are equipped with a device called the language acquisition device
(LAD) and universal grammar (UG) which accounts for the swift mastery of language among children despite
the extremely abstract nature of language. The Interactionists[9] believes that language is not a separate element
of the mind as language reflects the information gained through children's physical contact with the world.
Authors in [30] found that the existence of feedback had a significant impact on the structure of the network
used by learners to learn the properties of words in a natural language. A statistical learning system suggests that
learners track distributional information in their environment and use that information to derive the structure and
concepts they obtain about the sensory inputs. For example, in running speech, infants can segment words from
an artificial language by monitoring the transitional probability of syllables.
2.2 Second Language Acquisition
Acquisition of their vocabulary is a crucial part of learning a new language. Morphology in the linguistic sense
is the study of words, how they are created, and how they relate to other words in the same language. Study in
[31] discussed the neural signature in initial phase of morphological rule based learning of a novel language
(L2) in adults and suggests that even after a short exposure, adult language learners can acquire both novel
words and novel morphological rules of L2.
Bilingualism studies have identified ways in which a second language's neural representation (L2) varies from
that of the first language (L1) of an person[32]. In particular, there are many variations in activation between L
and L1, both in degree and magnitude. L2s tend not only to display more activity within traditional language
areas of the left hemisphere but also to enable more regions beyond the traditional language network. There are
two prevailing hypotheses about why L2 neural signatures vary from L1 signatures. The first is that, during L
learning, these variations reflect decreased neuroplasticity that occurs at a later age than learning with L1. L
learning needs increased neural capital on this account due to maturational changes in neural plasticity within
regions and pathways that enable first language learning [33]. The alternative hypothesis is that neural variations
in L1 versus L2 are caused instead by the fact that the L2 of individuals is typically lower in ability than their
L1. Therefore, the processing of L2 requires increased computational requirements and thus increased neural
resources[34]. The experimental results indicate that ability and AoA describe different functional and structural
networks within the bilingual brain, which we interpret as indicating distinct plasticity forms for age-dependent
effects versus experience and/or skills.
Authors in [35] consider structural changes to brain areas believed to support language roles during learning of a
foreign language. Experimental findings show that the volume of the hippocampus and the cortical thickness of
the left middle frontal gyrus, inferior frontal gyrus and superior temporal gyrus increase for interpreters
compared to controls. In interpreters with higher foreign language abilities, the right hippocampus and the left
superior temporal gyrus were both structurally more maleable[36].
Study in[37] investigated how the age at which L2 was acquired influenced brain structures in bilingual people.
This shows that AoA, language skills and current exposure rates are equally important in taking into account the
systemic differences. Structural changes related to bilingualism and multilingualism have also been reported,
bilinguals tend to have increased grey matter volume/density in Heschl’s gyrus [38], the left caudate [39] and
the left inferior parietal structure [40].
Authors in [41] explored the correlation between instructed second language acquisition (ISLA) skills and
identified a clear connection between attitude towards language learning and second language skills. The
analysis of language learning achievements in monozygotic and dizygotic twins [42] point to the possibility that
having a positive attitude towards language learning and the language class is related to how well students do in
ISLA independent of natural language abilities, teacher skill and L1-L2 relations.
In [43], authors examined the neural substrates of novel grammar learning in a group of healthy adults
conducted an experiment and study based on fMRI that, in terms of functional connectivity, the involvement of
the brain network during grammar acquisition is coupled with one's language learning ability.
2.3 Sign Language Acquisition
Children born deaf can not understand the languages spoken around them, and there is inadequate phonetic
information provided by the visual signal of speech to facilitate spontaneous language acquisition. For many of
these youngsters, language learning continues far beyond infancy after exposure to and immersion in a sign
language at ages. Variation in the period of language acquisition in the adult brain influences language
processing[44], [45]. In the classical language areas of the left hemisphere LH, fMRI studies of deaf native
signers have find activation with a trend towards bilateral activation of the frontal and temporal lobes. These
findings were observed using different tasks and triggers for distinct sign language namely American, British
and Japanese[46]–[48].
The learning age is linearly and inversely related to activation rates in anterior language regions and positively
related to activation rates in subsequent visual regions for linguistic tasks of American sign language (ASL)
sentences, grammatical judgment and phonemic hand judgment [29].
2.4 Skill Acquisition
Authors in [49] addressed that the expression of this neuroplasticity depends on the age at which learning starts
in several domains of skill acquisition. In studies aimed at determining the relationship between age of maturity
and brain plasticity, the fact that most abilities are learned late in childhood or adulthood has proven to be a
limit. According to [50], early sensory experiences tend to have the greatest capacity to improve neuronal
circuitry in the early years of development, When the brain is in active building up phase. Neuroimaging studies
of language development concentrate on the variations between simultaneous and concurrent bilinguals in brain
structure and function, and whether bilingualism is accomplished later in life. It also discusses the idea of an
optimal time in the production of languages and thus gives the relationship between the acquisition era and the
ultimate results[51].
Santiago Ramon Y Cajal (Nobel prize winner) in 1894, proposed that mental activity might induced
morphological changes in brain structure. Authors in[52] determined that the human brain structure expands and
get renormalized during skill acquisition. It is known as the expansion-renormalization model, according to
which neural processes related to learning always adopt a sequence of expansion, selection, and renormalisation.
[53]. The model foretell an initial increase in the density of grey matter, theoretically representing the growth of
neural capital such as neurons, synapses and glial cells, Accompanied by a selection process operating on this
new tissue which results in a complete or partial return to the baseline overall volume after selection has been
completed. To date, improvements in brain structure have been reported on different time scales, such as several
months of juggling training, medical examination study, space navigation training, learning of foreign
languages, etc.
For any language learning, the age of its acquisition matters a lot. The literature shows the importance
of age for learning a language, early language acquisition improves the probability of being proficient in a
language.For first language learning, social environment of infants also plays a significant role, age of learning,
nature of input language and teaching strategy is also important. Second language acquisition becomes easier if
it is learned in early age (before puberty)because during this period brain have more plasticity and it also have
lot of idea about language learning which is experienced during first language learning. Vocabulary and
grammar learning of second language is easier if it is done simultaneously or sequentially of L1 in early
childhood. Sign language acquisition is done in later age than infants, as it is generally learned by born deaf
children. Age of sign language learning also affects its proficiency. Skill development or expertise learning is
also depending on age and the language proficiency before getting that skill. Learning at later stage can be
improved by doing morphological learning.
Table 2: Review of Language acquisition in brain
S.N. Author Task Computation Method
Data Acquisition Method
Result
Language Acquisition
1 P. K. Kuhl et. al., 2010 [54]
Language and pre-reading in year two, third and fifth years.
Alpha, beta and gamma rhythms analysis.
EEG/ERPs/MEG/f MRI/NIRS
Early mastery of the phonetic units of language demands social learning.
2 R. I. Mayberry et. al., 2011[29]
American Sign Language, grammatical judgment and phonemic-hand judgment
t-statistics fMRI data The left lateralised activation pattern was observed
3 I. Kovelman et. al., 2012 [14]
Language task and Rhythm Task
t-test analysis functional Near Infrared (fNIRS) imaging
The right hemisphere overall displayed greater activation against the sluggish rhythmic stimulation, and the left hemisphere displayed greater activation compared to the quicker and slower frequencies.
4 J. Martensson et. al., 2012 [35]
Three months of intense foreign language studies
t-test on cortical thickness
MRI Structural changes in brain areas known for performing language roles during the learning of foreign languages. 5 S. Penicaud et. al., 2013 [28]
American Sign Language (ASL) voxel-based whole-brain correlational analysis.
fMRI Not only the functional but also the structural structure of the brain is impaired by lack of early language experience. 6 Miao Wei et. al., 2015 [37]
Language history questionnaire task
Cluster size, t- score.
MRI/fMRI/ PET In the right parietal cortex, earlier second-language sensitivity is correlated with greater volumes. Consistently, as AoA decreased, the cortical region of the right superior
considered essential to human language communication: the area of Wernicke and the area of Broca. The
accurate fasciculus is the brain region between the Wernicke region and the Broca area which connects the two
via bundles of nerve fibers. This part of the brain acts as a hub of transportation between the two areas mainly
concerned with speech and communication.
Comprehension of sentences depends crucially on deciding the thematic relationship between noun phrases, i.e.
defining who is doing what to whom. Study in [55] based on fMRI evaluated the relevant grammar and a key
factor underlying the assessed output in the verbal working memory. Voxel-based gray matter morphometry
showed that while the capacity of children to assign thematic roles in the left inferior temporal gyrus and the left
inferior frontal gyrus in positively correlated with gray matter likelihood (GMP). The verbal work memory-
related output in the left parietal operculum is positively associated with GMP extending into the posterior
superior temporal gyrus. Those areas are known to be involved in dynamic sentence processing in a particular
way. Results indicate a common GMP relationship in language-relevant brain regions and differential cognitive
abilities that direct their interpretation of the sentence.
EEG mu rhythms recorded at fronto-central electrodes are commonly considered to be measures of human
motor cortical activity as they are modulated when the participants perform an action, experience another's
action or even imagine an action. Study in [56] recorded the modulation of mu rhythms in time frequency (TF),
while participants interpreted the language of motion, abstract language and perceptive language. The findings
indicate that mu repression is correlated with the language of practice rather than with abstract and perceptive
language at fronto- locations. It also indicates that the activation takes place online through multiple words in
the sentence, based on semantic integration.
During sentence processing region of the left upper temporal sulcus, inferior frontal gyrus and left basal ganglia
show a systemic increase in brain activity as a function of constituent size, indicating their participation in
computing syntactic and semantic structures. Experiments in [57] for non-spoken sign language on deaf
participants show that the same network of language areas was found, while reading and sign language
processing created similar effects of the linguistic structure in the basal ganglia, the effect of structure in all
cortical language areas was greater for written language relative to sign language.
Based on evidence from neuroimaging, literature[58] reported both substantial overlap and unique linguistic
cortical activation between comprehension of the sign language and observation of gestural behavior. In the
upper / lower parietal lobe and the fusiform gyrus, overlaps in cortical activation are primarily observed.
Authors in[59]found that American Sign Language (ASL) stimulated more strongly the left inferior frontal
gyrus (IFG) and the middle superior temporal gyrus (STG) in deaf native signers than gestures expressing
roughly the same material. Here Graph Theoretical Analysis(GTA) is used on the neural dependent cognition
studies as an important complementary perspective to the activation research.
Study in [60] illustrates the semantic and grammatical processing of accented speech, both native and
international. Closer analysis of listeners who did not understand the foreign accent correctly indicated that
listeners who recognized the foreign accent displayed ERP responses for both grammatical and semantic errors.
By comparison, listeners who did not correctly recognize the foreign accent gave no ERP responses to the
foreign accented condition's grammatical errors, but displayed a late negativity to semantic errors.
Study in [61] indicates that mechanisms of the right hemisphere in the brain are essential to triggering elements
of event information that breach the linguistic meaning. The brain stimulates components of event-knowledge
that are semantically anomalous in context during learning.
In [62] authors propose that the prosodic information available during spoken language comprehension supports
the generation of online predictions for upcoming words, and that comprehension of spoken language during
serial visual presentation (SVP) reading, at least for quantifier sentences, may proceed more incrementally than
understanding. The analysis demonstrates that the comprehension of spoken sentence continues fully
incrementally, the results of truth meaning in both positive and negative quantifier sentences are alike. This also
suggests that people use the spoken language more effectively than written SVP feedback to produce online
predictions about coming words. During listening to natural speech, learning usually continues more
incrementally than during an ERP experiment with N400 results during SVP hearing.
Authors in [63] note that, during late childhood and adolescence, the cortical depiction of language
comprehension is added concentrated within the superior and middle temporal regions. Higher language ability
are correlated with greater right hemispheric engagement during the listening of stories. Language
comprehension is expressed more bilaterally than language output and a hemispheric dissociation with the
development of the left hemispheric language, but comprehension of the bilateral or right hemispheric language
is not uncommon even in healthy right handed subjects.
In [64], authors conducted an experiment and found that medial parietal lobe requires the production of
referential words. Analysis of the experiment based on fMRI is done using a pairwise t test of total cluster
activation which verified that each referential sub-condition was correlated with more activation than the non-
reference condition.
The prefrontal brain regions historically associated with language comprehension are the Wernicke area and the
Broca area.11 subjects of the Curtiss- Yamada Comprehensive Language Evaluation Receptive (CYCLE-R) are
taken to perform voxel based lesion symptom mapping (VLSM) based analysis of functional neuroimaging data
indicated that lesions to five left hemisphere brain regions affected performance on the CYCLE-R, including the
posterior middle temporal gyrus and underlying white matter, the anterior superior temporal gyrus, the superior
temporal sulcus and angular gyrus, mid frontal cortex in BA 46 and BA 47 of the inferior frontal gyrus. Analysis
also suggested that the middle temporal gyrus may be more important for comprehension at the word level,
while other regions may play a greater role at the level of the sentence.
3.1 Bilingualism: A large portion of the world's inhabitants is bilingual, and is flawlessly in over one language.
A bilingual speaker routinely produces and understands without difficulty sentences which belong to two (or
more) languages. Hence, knowing hoe two languages coexists in one brain with little disagreement or intrusion
in both codes is a theoretical and applied question of great interest. There is ongoing debate about whether early
and/or prolonged exposure to more than one language may lead to changes in patterns of brain activity during
language processing.
Authors in [65] performed an experiment involving highly qualified bilingual Spanish / Catalan and Spanish
monolinguals made grammatical and semantic decisions in Spanish while being tested for fMRI.Grammatical
judgement showed increased activation in IFG (BA 45), fusiform gyrus (BA37), occipital lobe (BA 18) and in
superior parietal lobe (SPL, BA 7). For monolingual group cortical activations were found in IFG (BA 45/46/9),
SFG (BA 6), BA 8/32and BA 18/23/37). Study indicates bilinguals are attracting new areas of the brain.
However, these different areas that depend on the learning age, language use, task circumstances, type of
stimulus, cognitive / linguistic demands, and possibly the characteristics and relative similarities between the
languages the bilingual speakers speak.
It has been shown that the two languages of Bilingual are simultaneously involved during listening, reading and
speaking, even when only one language is specifically required. This parallel activation was shown to promote
lexical access and to interfere in bilingual comprehension with the language processing. Research has shown
that when bilinguals process visual words, they experience co-activation of language, and use inhibitory
regulation to overcome non-target language competition. Authors in [66]suggest that the degree of language co-
activation in bilingual spoken word comprehension is modulated by the amount of regular exposure to non-
target language; and that bilinguals less affected by cross-language activation may also be more effective in
suppressing non-linguistic task intervention.
Findings in [67] showed that language processing can be considered as the result of a network of brain regions
interacting, rather than finding just a few brain areas to be involved in it. The experiment based on fMRI and its
interpretation showed the activation of BA 44 and BA45 to be left lateralized in the three tasks (receptive
semantic expressive paradigm), indicating roles in language phonology and semantic; however, their right
homologous areas were also involved, which may be due to their involvement in executive function, attention or
memory manipulation. On the contrary, BA 22 activation dominated at the right. The authors propose that right
BA22's contribution to language acquisition is an integral part of a broader chain comprising left IFG, bilateral
STG and lower parietal lobule. There are also studies that consider the right hemisphere as the seat for the
transmission of phonology and semance.
EEG-based studies in[68]have taken on two tasks: the semantic decision-making task and the task of reading.
The numerous experiment 1 wave maps indicate there was a frontal distribution of the disparity between literal
and novel metaphoric sentences. For both studies, the amplitudes of late positive complex (LPC) for novel
metaphoric sentences were decreased compared to those for anomalous sentences over parietal sites. While this
effect was clearly lateralized in experiment 2, in experiment 1 it posed a wider parietal distribution.
Authors in [69] performed an experiment focused on repeated transcranial magnetic stimulation (rTMS) taking
lexical decisions against basic tasks of judging. Findings provide evidence of an early motor cortex- TMS
intervention protocol creates a lateralized left, task and meaning contextual improvement in response latencies,
slowing down action-related word processing compared to faster abstract word reactions. The findings clearly
suggest causal involvement of different modality circuits in language understanding, suggesting that cognitive
phenomena of high order are based on simple biological mechanisms.
In [70], authors conducted a Near Infrared Spectroscopy (NIRS) experiment using the method of listening to
English sentences with six separate speeches. The findings showed that Japanese subjects had understood
speech with some of the characteristics of speech when amplitudes were expanded at certain frequency ranges.
The NIRS measurement also revealed that the enhancement of high frequency amplitudes ranging from 7000 to
8500 Hz increased concentration of Oxy-Hb in most language areas (BA 45/44/22).
Study in[71] shows that the neural representation of sentences in two languages is normal. From a mapping built
in English, the proposed model successfully predicted Portuguese sentences using brain positions and weights
applied to neutrally plausible semant features (NPSF). The mapping between the neural activation patterns and
analysis 7 R. Metusalem et. Al., 2016[61]
Expected, Event-Related, Event- Unrelated words and comprehension question answers are used
Several statistical analyses were conducted on mean ERP voltage measures
EEG Foster our understanding of the neural basis of event information activation and advance our understanding of how event awareness is activated in incremental understanding during creation of perceptions and elaborate inferences more generally. 8 D. Freunberger et. al., 2016[62]
N400 event-related potentials (ERP)
Linear Mixed Effects (LME) Models Using S Classes.
EEG People use the spoken language more effectively than written SVP feedback to produce online predictions of coming words. 9 Y. Yang et. al., 2017 [71]
English and Portuguese language reading
BOLD activation analysis
fMRI Proven ability to predict meta- language through cultures, people and bilingual status. 10 S. Grey et. al., 2017 [60]
Foreign-accented and native- accented speech
Mean ERP amplitudes
EEG/ERP Provide novel insights into understanding the impact of listener familiarity and foreign-emphasized speaker status on language processing neural correlates. 11 L. Liu et.al., 2017 [58]
Learning sign language by signers and Non-signers 'understanding of sign language.
Graph theoretical analysis (GTA)
fMRI When observing sign language, the hearing signers and non-signers showed identical cortical activations. The frequently activated network was structured differently between the two classes, however. 12 C. Brodbeck et. al., 2017[77]
Visuo-spatial referential Domains t-tests MEG/EEG Reports the medial parietal lobe participates in the production of referential words. 13 P. Chen et. al., 2017 [66]
Word pairs consisting of an English-Korean inter-lingual homophone
ANOVAs with relatedness
Event-related potential (ERP)
The amount of regular exposure to the non-target language modulates the degree of language co-activation in bilingual spoken word comprehension. 14 N. Vukovic et. al., 2017 [78]
Action words, abstract words and pseudo words
ANOVA, with the independent factors of Task
repetitive transcranial magnetic stimulation (rTMS)
Cortical motor regions play a vital role in understanding language.
15 K. Inada et. al., 2017 [70]
English speech task Enhanced amplitudes
Near-infrared spectroscopy system (NIRS)
English discourses with enhanced amplitudes within a certain frequency range can affect brain function activation in the language processing area and contribute to a better understanding of English speaking. 16 A. Moreno et.al., 2018 [57]
Sign language paradigm and written French stimuli
Z-score MRI/fMRI It suggests that the language network is systematically active in combinatorial language operations, comprising the left superior temporal sulcus, inferior frontal gyrus, and basal ganglia. 17 R. Alemi et. al., 2018 [67]
Word Production (WP) task, Auditory Responsive Naming (ARN) paradigm, Visual Semantic Decision (VSD) paradigm
Group ICA fMRI The language function should be regarded as the result of a network of brain regions collaborating.
18 K. Rataj et. al., 2018 [68]
Semantic decision and a reading task
t-test EEG The Late-Positive-Complex (LPC) pattern is modulated by both conventionality and task demand.
4Data Acquisition and Analysis Techniques
4.1 Data Acquisition
Over the last decade has shaped rapid developments in non-invasive practices that observe language processing
in human brain. They include Electroencephalography (EEG)/ Event-related Potentials (ERPs),
Magnetoencephalography (MEG), structural/resting-state Magnetic Resonance Imaging (rsMRI), functional
Magnetic Resonance Imaging (fMRI), Near Infrared Spectroscopy (NIRS), Diffusion Tensor Imaging (DTI),
Positron emission tomography (PET) etc.
4.1.1 Magnetic resonance imaging (MRI) can be paired with MEG and/or EEG, which offers static structural /
anatomical brain images. Structural MRIs display structural variations over the lifetime of brain regions and
they were recently used to predict second-language phonetic learning for adults. Magnetic resonance imaging
(MRI) can be paired with MEG and/or EEG, which offers static structural / anatomical brain images. Structural
MRIs display structural variations over the lifetime of brain regions and have recently been used to predict
phonetic learning of the second language of adults.[79]. In young children, structural MRI tests recognize the
size of different brain structures and these tests have been shown to be linked to language skills later in
childhood. When structural MRI images are superimposed on the physiological activity observed by MEG or
EEG, it is possible to enhance the spatial localization of brain activity reported by those methods. [80].
4.1.2 Functional magnetic resonance imaging (fMRI) is a common tool for human neuroimaging, since it
offers high spatial resolution maps of neural activity across the entire brain. [81]. The fMRI senses changes in
bloodoxygenation that happen in the neural activation response. Neural effects occur in milliseconds; but the
changes in bloodoxygenation they cause extend over many seconds, greatly restricting the temporal resolution
of fMRI. fMRI learning let exact location of brain activity and some groundbreaking study illustrate remarkable
similarities in the language-responsive structures in infants and adults. [82], [83].
4.1.3 Electroencephalography (EEG) is an electrophysiologic monitoring technique designed to capture brain
electrical activity. This is characteristically non-intrusive, with the electrodes located around the scalp, but, as in
electrocorticography, intrusive electrodes are sometimes used. EEG tests changes in voltage arising from ionic
current inside brain neurons. In clinical contexts, EEG mentions the monitoring over a period of time of the
normal electrical activity of the brain, as reported from multiple electrodes mounted on the scalp[84].
4.1.4 Event-related Potentials (ERPs) have been commonly used in infants and young children to study speech
and language production. ERPs, a part of the EEG, represent electrical activity that is time-locked to present a
particular sensory stimulus (for example, syllables or words) or a cognitive process (recognition within a
sentence or phrase of a semantic violation)[85]. By placing sensors on a child's scalp, it is possible to quantify
the behavior of neural networks firing in a synchronized and synchronous manner in open field environments,
and to detect voltage shifts that occur as a result of cortical neural activity[86],[87].
4.1.5 Magnetoencephalography (MEG) Is another method for brain imaging which tracks exquisite temporal
resolution of brain activity. The SQUID sensors positioned within the MEG helmet evaluate the minute
magnetic fields associated with electrical currents generated by the brain while performing sensory, motor, or
cognitive tasks. MEG facilitates the exact location of the neural currents accountable for magnetic field
sources[88],[89] the use of modern head monitoring methods and MEG to illustrate phonetic recognition in
newborns and infants in their first year of life..
4.1.6 Near-Infrared Spectroscopy (NIRS) Cerebral hemodynamic responses to neuronal activity are also
measured, but light absorption sensitive to haemoglobin concentration is used to assess activation [90]. NIRS
monitors increases in concentrations of blood oxy- and deoxy-haemoglobin in the brain, as well as increases in
total blood volume in various areas of the cerebral cortex using near-infrared light. The NIRS system can assess
activity in different brain regions by constantly monitoring the amount of haemoglobin in blood. In the first two
years of life, studies have started to surface on children, testing infant responses to phonemes as well as longer
periods of speech such as "motherese" and forward versus reversed sentences.
4.1.7 Diffusion Tensor Imaging (DTI) is an MRI-based neuroimaging technique that allows an estimation of
the position, orientation, and anisotropy of the white matter tracts of the brain [91].
4.1.8 Positron emission tomography (PET) tests pollutants from metabolically active chemicals injected into the
bloodstream, which are radioactively labeled. The emission data is processed by a computer to generate multi-
dimensional images of the distribution of the chemicals around the brain [92].
4.2 Data Analysis
Functional magnetic resonance imaging (fMRI) is a safe and non-invasive way of measuring brain function by
using brain activity-related signal changes. The method has become an omnipresent instrument of fundamental,
clinical, and cognitive neuroscience. This approach will calculate little changes in metabolism occurring in the
active part of the brain. We analyze the fMRI data in order to identify the parts of the brain that are involved in a
function, or to determine the changes that occur due to brain lesion in brain activities.
4.2.1 Statistical Analysis Methods
The efficiency of the fMR images is enhanced during the preprocessing stages. Thereafter, statistical analysis is
attempted to establish which voxels the stimulus stimulates. Many of the fMRI studies are focused on the
association between the hemodynamic response process and stimulation. Activation determines the changes in
the images to local severity. These methods can be divided into two specific categories: univariate methods
(methods for testing hypotheses), and multivariate methods (methods of exploration).
References
[1] A. D. Friederici, “The Brain Basis of Language Processing: From Structure to Function,” Physiol. Rev. , vol. 91, no. 4, pp. 1357– 1392, 2011.
[2] K. L. Sakai, “Language acquisition and brain development,” Science (80-. ). , vol. 310, no. 5749, pp. 815–819, 2005.
[3] E. H. Lenneberg, The biological foundations of language: 154-155. 1967.
[4] E. L. Newport and N. York, “Language development, critical periods in,” Encycl. Cogn. Sci. , pp. 737–740, 2003.
[5] J. R. Hurford, “The Evolution of the Critical Period for Language-Acquisition,” Cognition , vol. 40, no. 3, pp. 159–201, 1991.
[6] J. Gallaso, “First and second language acquisition,” 2003.. Gallaso, J. (2003). First and second language
acquisition. Retrieved from http://www.csun.edu/~galasso/lang1.htm.
[7] B. F. Skinner, “Cognitive science and behaviourism,” British Journal of Psychology , vol. 76, no. 3. pp. 291–301, 1985.
[8] N. Chomsky, “Verbal behaviour,” Language (Baltim). , vol. 35, no. 1, pp. 26–58, 1959.
[9] J. Pascual-Leone, “Vygotsky, Piaget and the problems of Plato,” Swiss J. Psychol. , vol. 55, no. 2, pp. 19–31, 1996.
[10] P. M. Lightbown and N. Spada, How languages are learned. Oxford: Oxford University Press, 2006.
[11] M. Walton, D. Dewey, and C. Lebel, “Brain white matter structure and language ability in preschool-aged children,” Brain Lang. , vol. 176, no. September 2016, pp. 19–25, 2018.
[12] J. M. Carroll, B. Maughan, R. Goodman, and H. Meltzer, “Literacy difficulties and psychiatric disorders: Evidence for comorbidity,” J. Child Psychol. Psychiatry Allied Discip. , vol. 46, no. 5, pp. 524–532, 2005.
[13] M. Smits, L. C. Jiskoot, and J. M. Papma, “White Matter Tracts of Speech and Language.,” Semin. Ultrasound CT MRI , vol. 35, no. 5, pp. 504–516, 2014.
[14] I. Kovelman, K. Mascho, L. Millott, A. Mastic, B. Moiseff, and M. H.Shalinsky, “At the rhythm of language: Brain bases of language-related frequency perception in children,” Neuroimage , vol. 60, no. 1, pp. 673–682, 2012.
[15] U. Goswami, “A temporal sampling framework for developmental dyslexia,” Trends Cogn. Sci. , vol. 15, pp. 3–10, 2011.
[16] L. A. Petitto, How the brain begets language: on the neural tissue underlying human language acquisition. Cambridge, UK: The Cambridge Companion to Chomsky. Cambridge University Press, 2005.
[17] A. L. Giraud, A. Kleinschmidt, D. Poeppel, T. E. Lund, R. S. Frackowiak, and H. Laufs, “Endogenous cortical rhythms determine cerebral specialization for speech perception and production,” Neuron , vol. 56, pp. 1127–1134, 2007.
[18] H. Luo, Z. Liu, and D. Poeppel, “Auditory cortex tracks both auditory and visual stimulus dynamics using low-frequency neuronal phase modulation,” PLoS Biol , vol. 8, 2010.
[19] B. Morillon, “Neurophysiological origin of human brain asymmetry for speech and language,” Pnas. 2010.
[20] A. Shusterman and P. Li, “Frames of reference in spatial language acquisition,” Cogn. Psychol. , vol. 88, pp. 115–161, 2016.
[21] S. C. Levinson, “Frames of reference and Molyneux’s question: Cross-linguistic evidence,” in Language and space , P. Bloom, L. N. Peterson, and M. F. Garrett, Eds. Cambridge: MIT Press, 1996, pp. 109–169.
[22] S. C. Levinson, Space in language and cognition: Explorations in cognitive diversity. Cambridge, UK: Cambridge University Pree, 2003.
[23] A. Majid, M. Bowerman, S. Kita, D. Haun, and S. Levinson, “Can language restructure cognition? The case for space,” Trends Cogn. Sci. , vol. 8, no. 3, pp. 108–114, 2004.
[24] E. Pederson, E. Danziger, D. Wilkins, S. Levinson, S. Kita, and G. Senft, “Semantic typology and spatial conceptualization,” Language (Baltim). , vol. 74, no. 3, pp. 557–589, 1998.
[25] D. B. M. Haun, C. J. Rapold, J. Call, G. Janzen, and S. C. Levinson, “Cognitive cladistics and cultural override in Hominid spatial cognition,” in Proceedings of the National Academy of Sciences of the United States of America , 2006, pp. 17568–17573.
[26] J. Pyers, A. Shusterman, A. Senghas, E. A. Spelke, and K. Emmorey, “Evidence from users of an emerging sign language reveals that language supports spatial cognition,” in Proceedings of the National Academy of Sciences , 2010, pp. 12116–12120.
[27] E. Partanen, A. Leminen, S. de Paoli, A. Bundgaard, O. S. Kingo, P. Krøjgaard, and Y. Shtyrov, “Flexible, rapid and automatic neocortical word form acquisition mechanism in children as revealed by neuromagnetic brain response dynamics,” Neuroimage , vol. 155, no. April, pp. 450–459, 2017.
[28] S. Pénicaud, D. Klein, R. J. Zatorre, J. K. Chen, P. Witcher, K. Hyde, and R. I. Mayberry, “Structural brain changes linked to delayed first language acquisition in congenitally deaf individuals,” Neuroimage , vol. 66, pp. 42–49, 2013.
[29] R. I. Mayberry, J. K. Chen, P. Witcher, and D. Klein, “Age of acquisition effects on the functional organization of language in the adult brain,” Brain Lang. , vol. 119, no. 1, pp. 16–29, 2011.
[30] E. Plante, D. Patterson, R. Gómez, K. R. Almryde, M. G. White, and A. E. Asbjørnsen, “The nature of the language input affects brain activation during learning from a natural language,” J. Neurolinguistics , vol. 36, pp. 17–34, 2015.
[31] V. Havas, M. Laine, and A. Rodríguez Fornells, “Brain signatures of early lexical and morphological learning of a new language,”
Neuropsychologia , vol. 101, no. April, pp. 47–56, 2017.
[32] E. S. Nichols and M. F. Joanisse, “Functional activity and white matter microstructure reveal the independent effects of age of acquisition and proficiency on second-language learning,” Neuroimage , vol. 143, pp. 15–25, 2016.
[33] P. Li, J. Legault, and K. A. Litcofsky, “Neuroplasticity as a function of second language learning : Anatomical changes in the human brain,” Cortex , vol. 58, pp. 301–324, 2014.
[34] J. A. Newman and A. Tremblay, “The Influence of Language Proficiency on Lexical Semantic Processing in Native and Late Learners of English,” J. Cogn. Neurosci. , vol. 24, no. 5, pp. 1205–1223, 2012.
[35] J. Mårtensson, J. Eriksson, N. C. Bodammer, M. Lindgren, M. Johansson, L. Nyberg, and M. Lövdén, “Growth of language- related brain areas after foreign language learning,” Neuroimage , vol. 63, no. 1, pp. 240–244, 2012.
[36] M. Stein, A. Federspiel, T. Koenig, M. Wirth, W. Strik, R. Wiest, D. Brandeis, and T. Dierks, “Structural plasticity in the language system related to increased second language proficiency,” Cortex , vol. 48, no. 4, pp. 458–465, 2012.
[37] M. Wei, A. A. Joshi, M. Zhang, L. Mei, F. R. Manis, Q. He, R. L. Beattie, G. Xue, D. W. Shattuck, R. M. Leahy, F. Xue, S. M. Houston, C. Chen, Q. Dong, and Z. L. Lu, “How age of acquisition influences brain architecture in bilinguals,” J. Neurolinguistics , vol. 36, pp. 35–55, 2015.
[38] V. Ressel, C. Pallier, N. Ventura-Campos, B. Diaz, A. Roessler, C. Avila, and N. Sebastian-Galles, “An Effect of Bilingualism on the Auditory Cortex,” J. Neurosci. , vol. 32, no. 47, pp. 16597–16601, 2012.
[39] L. Zou, G. Ding, J. Abutalebi, H. Shu, and D. Peng, “Structural plasticity of the left caudate in bimodal bilinguals,” Cortex , vol. 48, no. 9, pp. 1197–1206, 2012.
[40] P. A. Della Rosa, G. Videsott, V. M. Borsa, M. Canini, B. S. Weekes, R. Franceschini, and J. Abutalebi, “A neural interactive location for multilingual talent,” Cortex , vol. 49, no. 2, pp. 605–608, 2013.
[41] I. Antón-Méndez, E. M. Ellis, W. Coventry, B. Byrne, and V. H. P. van Daal, “Markers of success: A study of twins’ instructed second language acquisition,” Learn. Individ. Differ. , vol. 42, pp. 44–52, 2015.
[42] M. Coventry, I. Anton-Mendez, E. M. Elis, C. Levisen, B. Byrne, and V. H. O. Van Daal, “The etiology of individual differences in second language acquisition in australian school students: A behavior-genetic study,” Lang. Learn. , vol. 62, no. 3, pp. 880–901,
[43] O. Kepinska, M. de Rover, J. Caspers, and N. O. Schiller, “Whole-brain functional connectivity during acquisition of novel grammar: Distinct functional networks depend on language learning abilities,” Behav. Brain Res. , vol. 320, pp. 333–346, 2017.
[44] R. I. Mayberry, “When timing is everything: Age of first language acquisition effects on second language learning,” Appl. Psycholinguist. , vol. 282007/12/, pp. 537–549, 2007.
[45] R. I. Mayberry, “Early Language acquisition and Adult Language Ability: What Sign Language Reveals about the Critical Period for Language,” in Deafness , vol. 2, 2009, pp. 175–176.
[46] J. Kassubek, G. Hickok, and P. Erhard, “Involvement of classical anterior and posterior language areas in sign language production, as investigated by 4 T functional magnetic resonance imaging,” Neurosci. Lett. , vol. 364, pp. 168–172, 2004.
[47] S. McCullough, K. Emmorey, and M. Sereno, “Neural Organization for recognition of grammatical and emotional facial expressions in deaf ASL signers and hearing nonsigners,” Cogn. Brain Res. , vol. 22, pp. 193–203, 2005.
[48] K. L. Sakai, Y. Tatsuno, K. Suzuki, H. Kimura, and Y. Ichida, “Sign and speech: Amodal commonality in left hemisphere dominance for comprehension of sentences,” Brain , vol. 128, pp. 1407–1417, 2005.
[49] J. A. Berken, V. L. Gracco, and D. Klein, “Early bilingualism, language attainment, and brain development,” Neuropsychologia , vol. 98, no. September 2016, pp. 220–227, 2017.
[50] M. Butz, I. D. Steenbuck, and A. Van Ooyen, “Homeostatic structural plasticity increases the efficiency of small-world networks,” Front. Synaptic Neurosci. , vol. 6, no. APR, pp. 1–14, 2014.
[51] G. Li, J. Nie, L. Wang, F. Shi, W. Lin, J. H. Gilmore, and D. Shen, “Mapping region-specific longitudinal cortical surface expansion from birth to 2 years of age,” Cereb. Cortex , vol. 23, no. 11, pp. 2724–2733, 2013.
[52] E. Wenger, C. Brozzoli, U. Lindenberger, and M. Lövdén, “Expansion and Renormalization of Human Brain Structure During Skill Acquisition,” Trends Cogn. Sci. , vol. 21, no. 12, pp. 930–939, 2017.
[53] H. Makino, E. J. Hwang, N. G. Hedrick, and T. Komiyama, “Circuit Mechanisms of Sensorimotor Learning,” Neuron , vol. 92, no. 4, pp. 705–721, 2016.
[54] P. K. Kuhl, “Brain Mechanisms in Early Language Acquisition,” Neuron , vol. 67, no. 5, pp. 713–727, 2010.
[55] A. Fengler, L. Meyer, and A. D. Friederici, “Brain structural correlates of complex sentence comprehension in children,” Dev. Cogn. Neurosci. , vol. 15, pp. 48–57, 2015.
[56] I. Moreno, M. de Vega, I. León, M. Bastiaansen, A. Glen Lewis, and L. Magyari, “Brain dynamics in the comprehension of action-related language. A time-frequency analysis of mu rhythms,” Neuroimage , vol. 109, pp. 50–62, 2015.
[57] A. Moreno, F. Limousin, S. Dehaene, and C. Pallier, “Brain correlates of constituent structure in sign language comprehension,” Neuroimage , vol. 167, no. April 2017, pp. 151–161, 2018.
[58] L. Liu, X. Yan, J. Liu, M. Xia, C. Lu, K. Emmorey, M. Chu, and G. Ding, “Graph theoretical analysis of functional network for comprehension of sign language,” Brain Res. , vol. 1671, pp. 55–66, 2017.
[59] A. J. Newman, T. Supalla, N. Fernandez, E. L. Newport, and D. Bavelier, “Neural systems supporting linguistic structure ,
of auditory discrimination,” Exp. Neurol. , vol. 190, no. SUPPL. 1, pp. 44–51, 2004.
[89] T. Imada, Y. Zhang, M. Cheour, S. Taulu, A. Ahonen, and P. K. Kuhl, “Infant speech perception activates Broca’s area: A developmental magnetoencephalography study,” Neuroreport , vol. 17, no. 10, pp. 957–962, 2006.
[90] R. N. Aslin, “Near-infrared spectroscopy for functional studies of brain activity in human infants : promise , prospects , and challenges,” vol. 10, no. February, pp. 10–12, 2005.
[91] D. Le Bihan, C. Poupon, C. A. Clark, S. Pappata, N. Molko, and H. Chabriat, “Diffusion Tensor Imaging : Concepts and Applications,” vol. 546, pp. 534–546, 2001.
[92] D. L. Bailey, W. David, P. E. Valk, M. N. Maisey, and A. Greenspan, “Positron Emission Tomography : Basic Orthopedic Imaging : A Practical,” vol. 241, no. 1, pp. 45–46, 2006.
[93] K. J. Friston Holmes, A.P., Worsley, K.J., Poline, J.P., Frith, C.D., Frackowiak, “Statistical parametric maps in functional imaging: a general linear approach.,” Hum. Brain Mapp. , vol. 2, 1995.
[94] B. Y. D. B. Rowe and R. G. Hoffmann, “Analysis in fMRI,” no. April, 2006.
[95] L. I. Kuncheva and J. J. Rodríguez, “Classifier ensembles for f MRI data analysis : an experiment,” vol. 28, pp. 583–593, 2010.
[96] D. W. Shattuck, D. W. Shattuck, R. M. Leahy, and R. M. Leahy, “BrainSuite: An automated cortical surface identi cation tool,” Methods , vol. 6, pp. 129–142, 2002.
[97] A. Fedorov, R. Beichel, J. Kalphaty-Cramer, J. Finet, J.-C. Fillion-Robbin, S. Pujol, C. Bauer, D. Jennings, F. Fennessy, M. Sonka, J. Buatti, S. Aylward, J. V. Miller, S. Pieper, and R. Kikinis, “3D slicers as an image computing platform for thw quantitative imaging network,” Magn. Reson. Imaging , vol. 30, no. 9, pp. 1323–1341, 2012.
[98] “AFNI.”.
[99] S. Whitfield-Gabrieli and A. Nieto-Castanon, “ Conn : A Functional Connectivity Toolbox for Correlated and Anticorrelated Brain Networks,” Brain Connect. , vol. 2, no. 3, pp. 125–141, 2012.
[100] L. Umr and I. Université, “EEGNET manual.”
[101] “FreeSurfer.”.
[102] W. T. C. for Neuroimaging, “SPM - Statistical Parametric Mapping.”.
[103] FMRIB Analysis Group, “FSL - FslWiki.” 2012.
[104] “NITRC.”.
