ReviewMeta-analyzing left hemisphere language areas: Phonology, semantics, and sentence processing
Introduction
In the nineteenth and twentieth centuries, our understanding of how the human brain analyzes and produces language was shaped by aphasiology. This approach helped to define a model of language architecture, in which Broca's and Wernicke's areas—two fairly large and loosely defined cortical areas located in the left frontal and temporal lobes—were assigned the leading roles in language production and comprehension, respectively (Geschwind, 1970).
In the 1980s, the advent of non-invasive functional brain imaging techniques, such as positron emission tomography (PET) and functional magnetic resonance imaging (fMRI), shifted the emphasis towards investigating the neural bases of language in the normal human brain. These techniques have enabled the investigation of specific components of brain language implementation through the tailoring of experimental neuro-linguistic paradigms. Fifteen years have passed since the initial pioneering studies (Petersen et al., 1988, Zatorre et al., 1992), and the amount of information coming from functional imaging studies offers the opportunity to propose an updated view on brain–language relationships.
Starting from 260 articles published between 1992 and 2004, the present meta-analysis is constituted with 129 scientific reports, including 262 component-specific cognitive contrasts reporting 730 activation peaks in the left hemisphere cortex. These contrasts were selected from studies investigating three language-processing component classes, namely, “phonological,” “semantic,” and “sentence processing” components. This classification emerges not from language models per se but more pragmatically from the design of the selected functional imaging protocols, which compared cognitive conditions with increasingly complex verbal material.
Selection of articles and contrasts to be included in the present meta-analysis was based upon five major criteria: 1—we limited the investigation to studies on normal volunteers; 2—we retained those that reported their results in stereotactic coordinates, thereby excluding articles using a region-of-interest (ROI) approach; 3—we selected component-specific contrasts, thereby excluding studies that used low-level reference tasks, such as cross-fixation or rest, for semantic and sentence processing investigations; 4—we selected activation peaks located in the left–middle and inferior frontal gyrus, including the insula, the superior, middle, and inferior temporal gyri and inferior parietal gyrus; 5—peaks that felt outside the brain or in the white matter were thus excluded. The selected peaks for each class were then submitted to a spatial clustering algorithm in the standard neuroanatomical reference space shared by the brain mapping community (i.e., Montreal Neurological Institute (MNI) space), after appropriate conversion when necessary.
The present approach is in line with Poldrack's meta-analysis of the inferior frontal gyrus that allowed an anatomo-functional segregation for phonological and semantic processing (Poldrack et al., 1999). It is based on the whole brain three-dimensional and quantitative methodology that our laboratory has developed for the identification of the different pathways for word reading (Jobard et al., 2003). This approach appears complementary to recent reviews on language that investigated one language component (word processing (Indefrey and Levelt, 2004, Price, 2000), semantic processing (Bookheimer, 2002), sentence processing (Friederici, 2002)) and confronted a reduced number of investigations to propose a synthesis of functional imaging results within a given theoretical frame. These latter reviews are based on a fine-grained analysis of the paradigms and provide important insights on the neural representation of language. However, they describe functional anatomy of language at the gyral level only, leaving open the question of a more accurate cortical localization of a given process and thus that of specificity and interactions of different processes at regional level. Complementary to these works, the present methodology provides a fine-grained analysis of the brain functional activations, searching for the existence of areas dedicated to each of the language components–for instance, the segregation of semantic and phonological areas within the inferior frontal gyrus–and/or the opportunity to question the role of nearby or overlapping cortical areas. Given its power in terms of number of subjects and range of language paradigms, our meta-analysis approach offers the possibility of identifying regions that could have been overlooked by single protocols and that would correspond to invariants of a given language component. We think that such an approach is the only way to provide empirical arguments to support the theoretical shift proposed by M. Mesulam and J. Fuster that cognition in the brain is supported by overlapping and interactive large-scale cognitive networks rather than modules (Fuster, 2003, Mesulam, 1990).
Section snippets
Contrast selection
To meta-analyze the language neuroimaging literature, a classification of language-processing components had to be adopted in order to sort the various cognitive states that were reported. We retained mainly component-specific contrasts, i.e., contrasts designed to target a precise component; low-level contrasts, such as comparisons between a language task with a resting state or a low level baseline, were discarded because such contrasts result in the activation of multiple peaks corresponding
Results
The clustering process for each language component led to the segregation of three to five clusters in the left frontal lobe and five to seven in the left temporal and parietal areas (Fig. 1, bottom). Cluster standard deviations ranged from 8 to 19.5 mm (Table 4). While similar numbers of frontal and temporal peaks were found for phonological contrasts (frontal: 125, temporal: 122, see Table 4), a larger number of temporal peaks were noted for semantic (frontal: 145, temporal: 177) and sentence
Methodological issues
The first point that has to be discussed concerns the consequences of the methodological constraints applied on the studies selection. It must be underlined that the restriction we applied prevented us to include some studies that provide key elements in the comprehension of language networks. This pitfall is inherent to any meta-analysis, and we tried to overcome this limitation by integrating such articles within the discussion.
The second point is related to the large-grained resolution in
Summary and conclusions
This exploratory approach of left hemisphere language areas provided a set of results that reinforces and refines hypotheses emerging from isolated functional imaging studies.
Above all is the involvement of an elementary audio–motor loop for phonological processing, allowing a motor-sound-based representation for language sounds and involved in whether language is heard or enunciated, as had been proposed by Buchsbaum for auditory areas (Buchsbaum et al., 2001) and by Wilson for motor areas (
Acknowledgments
The authors are deeply indebted to Franck Lamberton and Nicolas Delcroix for their help in data analysis and to an anonymous reviewer for the clarity and relevance of her (his) comments that led the authors to revise their interpretation of some results.
References (197)
- et al.
A comparison of neural circuits underlying auditory and visual object categorization
NeuroImage
(2002) - et al.
PET studies of memory: novel and practised free recall of complex narratives
NeuroImage
(1995) - et al.
Event-related fMRI reveals cortical sites involved in contextual sentence integration
NeuroImage
(2002) - et al.
Human temporal-lobe response to vocal sounds
Brain Res. Cogn. Brain Res.
(2002) - et al.
Neural correlates of syntactic movement: converging evidence from two fMRI experiments
NeuroImage
(2004) - et al.
The role of frontopolar cortex in subgoal processing during working memory
NeuroImage
(2002) - et al.
Unitary vs. multiple semantics: PET studies of word and picture processing
Brain Lang.
(2004) - et al.
Recognition of emotional prosody and verbal components of spoken language: an fMRI study
Cogn. Brain Res.
(2000) - et al.
Role of left posterior superior temporal gyrus in phonological processing for speech perception and production
Cogn. Sci.
(2001) - et al.
PET studies of syntactic processing with auditory sentence presentation
NeuroImage
(1999)
The effect of semantic category and knowledge type on lexical–semantic access: a PET study
NeuroImage
Virtual in vivo interactive dissection of white matter fasciculi in the human brain
NeuroImage
The Talairach coordinate of a point in the MNI space: how to interpret it
NeuroImage
Overlap and dissociation of semantic processing of Chinese characters, English words, and pictures: evidence from fMRI
NeuroImage
Sentence complexity and input modality effects in sentence comprehension: an fMRI study
NeuroImage
Neural correlates of naming actions and of naming spatial relations
NeuroImage
Form and content: dissociating syntax and semantics in sentence comprehension
Neuron
Neural responses to morphological, syntactic, and semantic properties of single words: an fMRI study
Brain Lang.
Differential activation of right and left posterior sylvian regions by semantic and phonological tasks: a positron-emission tomography study in normal human subjects
Neurosci. Lett.
The pursuit of brain–language relationships
Brain Lang.
Lesion analysis of the brain areas involved in language comprehension
Cognition
What does the frontomedian cortex contribute to language processing: coherence or theory of mind?
NeuroImage
Effects of lexicality, frequency, and spelling-to-sound consistency on the functional anatomy of reading
Neuron
Other minds in the brain: a functional imaging study of “theory of mind” in story comprehension
Cognition
Location-probability profiles for the mouth region of human primary motor–sensory cortex: model and validation
NeuroImage
Towards a neural basis for auditory sentence processing
TICS
Network memory
TINS
Linkage at the top
Neuron
Reading the mind in cartoons and stories: an fMRI study of ‘theory of mind’ in verbal and nonverbal tasks
Neuropsychologia
Dissociation of mechanisms underlying syllogistic reasoning
NeuroImage
Specialization in the left prefrontal cortex for sentence comprehension
Neuron
Topographic segregation and convergence of verbal, object, shape and spatial working memory in humans
Neurosci. Lett.
Broca's area in the human brain is involved in the selection of grammatical gender for language production: evidence from event-related functional magnetic resonance imaging
Neurosci. Lett.
Interindividual variation of cerebral activation during encoding and retrieval of words
Eur. Psychiatr.
Dorsal and ventral streams: a framework for understanding aspects of the functional anatomy of language
Cognition
From perception to sentence comprehension: the convergence of auditory and visual information of language in the left inferior frontal cortex
NeuroImage
Selective enhancement of functional connectivity in the left prefrontal cortex during sentence processing
NeuroImage
The effects of attention on speech perception: an fMRI study
Brain Lang.
The spatial and temporal signatures of word production components
Cognition
Brain areas engaged during visual judgments by involuntary access to novel semantic information
Vision Res.
Working memory
Science
The neural substrates for concrete, abstract, and emotional word lexica: a positron emission tomography
J. Cogn. Neurosci.
Adaptation to speaker's voice in right anterior temporal lobe
NeuroReport
Voice-selective areas in human auditory cortex
Nature
Function of the left planum temporale in auditory and linguistic processing
Brain
Human brain mapping language areas identified by functional magnetic resonance imaging
J. Neurosci.
Conceptual processing during the conscious resting state: a functional MRI study
J. Cogn. Neurosci.
Human temporal lobe activation by speech and nonspeech sounds
Cereb. Cortex
Neural correlates of lexical access during visual word recognition
J. Cogn. Neurosci.
Transcortical sensory aphasia: revisited and revised
Brain
Cited by (1380)
Corticocortical Evoked Potentials in Eloquent Brain Tumor Surgery. A Systematic Review
2024, World NeurosurgeryFunctional connectivity during morphosyntactic processing: An fMRI study in balanced Turkish-Persian bilinguals
2023, Journal of NeurolinguisticsComparative efficacy of non-invasive brain stimulation for post-stroke cognitive impairment: a network meta-analysis
2024, Aging Clinical and Experimental Research
- 1
These authors contributed equally to this work.