The following is our proposal for a special issue of Cognition submitted to Jacques Mehler in 2000 and then accepted after a couple of rounds of review in January 2001. The proposal was almost a mini paper itself! It finally appeared in print in the May 2004 issue.
Towards a New Functional Anatomy of Language
A Proposal for a Special Issue of Cognition
Gregory Hickok
Department of Cognitive Sciences
University of California, Irvine
David Poeppel
Department of Linguistics, and
Department of Biology
University of Maryland, College Park
Accepted 24 January 2001
There has been virtually no progress in the development of models of the functional anatomy of language since the late 19th century. Although this statement may appear to be an overstatement, it is difficult to refute. To be sure, there has been a tremendous accumulation of knowledge within the field since the first anatomical model of language was put forth by Wernicke in 1874: We now have a far better understanding (i) of the nature of language representation, (ii) of the stages and operations involved in processing language in real time, (iii) of the symptom complex of aphasia and the distribution of lesions linked to these various symptoms and syndromes, (iv) of the brain regions which are physiologically active while normal subjects perform various language tasks, and (v) of the structure and functional organization of the brain generally. But despite this accumulation of knowledge, there is no functional anatomical model of language which does a better job at capturing the range of empirical facts than Wernicke’s 1874 model, even with its many acknowledged shortcomings.
Put another way, the parallel development of linguistic and psycholinguistic facts and models on the one hand, and anatomical facts and models on the other, has remained just that, parallel. This observation is highlighted by a figure (reproduced below), from the language chapter in Gazzaniga et al.’s recent (1998) Cognitive Neuroscience textbook, which shows a version of Levelt’s psycholinguistic model and its (non-)relation to brain organization.
The primary goal of this special issue to push the field closer to bridging this gap. Explicit attempts at bridging this gap have already begun in the form of several recent publications by a number of authors, using a wide variety of cognitive neuroscientific methods. Several authors, for example, have made serious attempts to link portions of psycholinguistic models with the brain systems that support them (Blumstein, 1995; Boatman, et al., 1998; Indefrey & Levelt, 2000; Ullman, et al., 1997). Others have been instrumental in challenging misconceptions of the neuroanatomy of language (Anderson, et al., 1999; Dronkers, Redfern, & Knight, 2000; Hickok & Poeppel, 2000; Norris & Wise, 2000), and even identifying new regions involved in language processes (Damasio, Grabowski, Tranel, Hichwa, & Damasio, 1996; Dronkers, 1996). Still others have made an effort to reinterpret the functional anatomy of language in terms consistent with what is known about functional brain organization more generally (Aboitiz & García V., 1997; Hickok & Poeppel, 2000; Ullman, et al., 1997). It is time to bring these fresh perspectives together in one volume.
We have invited several of the leaders of this “new movement” to contribute a paper for a special issue of Cognition; all have accepted. Each author has been presented with the charge to write a paper which not only provides a competent review of the data within a sub-field of language-brain research, but attempts to place those facts in a theoretical context. This context may be a psycholinguistic theory, an existing or novel functional anatomical model, or ideally, both. This charge precludes reviews that are merely progress reports from the author’s lab, or lists of recent findings. Instead, the papers will generate new testable hypotheses, which will serve to guide future work.
The contributors represent a wide range of sub-domains of language-brain research – e.g., speech perception, lexical and morphological processes, syntactic processes, speech production, working memory -- as well as a wide range of methodological techniques – e.g., imaging, lesion, cortical stimulation, split brain, neurodegeneration. These authors were carefully chosen. Since our goal is to spark further development of integrative large-scale models of the functional anatomy of language, we chose contributors with complementary theoretical and methodological backgrounds. And while individual authors inevitably have specific expertise in, or bias towards, a particular method or theoretical approach, we have asked authors to adopt a broader perspective, both in data and theory, in the development of their contributions. One way we plan to do this is to encourage interaction among the contributors during the manuscript preparation process which will allow authors to draw on the resources and expertise of other members of the group. (For example, we have established a web-site accessible to contributors where we will post manuscript abstracts, outlines, drafts, and so on.) Our expectation is that together, this collection of papers will provide a scaffolding for the further development of new large scale models of the functional anatomy of language.
Outline of the Special Issue
The authors listed below comprise our final list of contributors. All have accepted our invitation to contribute a paper.
The first two papers (Wise & Scott, and Boatman) address the cortical organization of auditory perception of speech up to the word level. Wise & Scott start with a review of neuropsychological and physiological data to paint a general picture of the neural basis of speech perception and then draw on evidence from functional neuroimaging to further parcel out the subdivisions of auditory cortex. Wise & Scott’s emphasis will be on pre-lexical processes. Boatman addresses a similar issue with an increase in emphasis on higher-level issues concerning the interface between sub-lexical, lexical, and post-lexical processes in the perception/comprehension of speech. She will draw on her expertise with a wide range of lesion-based work (broadly construed to include stimulation mapping, split brain, Wada, and aphasia studies).
The next three papers (Dronkers et al; Ullman; and Friederici) deal with the neural basis of lexical and grammatical processes. Dronkers et al. will review word- and sentence-level comprehension deficits in aphasia and present two new large-scale lesion studies addressing this issue with special emphasis on how word- and sentence-level processes interact in language comprehension. Ullman will focus on the psycholinguistic distinction between lexical and grammatical processes, using data from a wide variety of sources to argue for a neural distinction between lexicon and grammar that parallels the distinction between explicit and procedural knowledge more generally. Friederici will review evidence for the recent suggestion that syntactic processes may be supported by anterior temporal lobe systems. She will also discuss possible roles for Broca’s area in syntactic processing, and how these temporal and frontal systems may interact.
The sixth and seventh papers (Levelt & Indefrey; and Damasio et al.) will tackle the issue of the anatomy of speech production. Levelt & Indefrey will attempt to make sense of a large number of functional imaging studies of speech production by decomposing the tasks into psycholinguistically relevant components. On the basis of this meta-analysis, and with supporting data from other methods, they will argue for a set of anatomical correlates of psycholinguistic models of speech production. Damasio et al. will present a major review of the neuroanatomy of lexical access in categorical naming along with the new experiments from their lab.
The final paper (Hickok & Poeppel) will propose a new model of the functional anatomy of language (including working memory) which integrates findings from psycholinguistics, aphasia, neuroimaging, and the surgical-based techniques. A central theme of the proposal is that the functional anatomy of language can be understood in terms of more general models of functional anatomy.
List of Contributors and Paper Titles
Introduction: Towards a New Functional Anatomy of Language
Gregory Hickok and David Poeppel
(10 pages)
Organisation of Cortical Systems for Speech Perception
Richard Wise and Sophie Scott
(30 pages)
Neural Bases of Auditory Language Processing: New Evidence from Lesion Studies
Dana Boatman
(15 pages)
Lesion Analysis of Word- and Sentence-Level Comprehension Deficits in Aphasia
Nina F. Dronkers, David P. Wilkins, Robert D. Van Valin Jr., Brenda B. Redfern and Jeri J. Jaeger
(40 pages)
The Contribution of Brain Memory Circuits to Language
Michael Ullman
(35 pages)
Neural Basis of Syntactic Processing
Angela Friederici
(25 pages)
The Spatial and Temporal Signatures of Language Production Components
Willem J.M. Levelt and Peter Indefrey
(30 pages)
Neural Correlates of Category Related Naming
H. Damasio, A. Damasio, D. Tranel, T. Grabowski and R. Adolphs
(50 pages)
Dorsal and Ventral Streams in Speech and Language Processes
Gregory Hickok and David Poeppel
(45 pages)
Gregory Hickok and David Poeppel
University of California, Irvine, and University of Maryland, College Park
Towards a New Functional Anatomy of Language
1.0 The Old Functional Anatomy of Language
1.1 What's right with the Wernicke-Lichtheim model
1.2 What's wrong with the Wernicke-Lichtheim model
2.0 The New Functional Anatomy of Language
2.1 Why we need it: there has been no viable replacement for Wernicke-Lichtheim
2.2 Why it's possible now: advances in neuroscience and technology
2.3 Why it hasn't yet emerged: lack of integration across neuroscience sub-fields and methods
2.4 What it should look like: constraints on model development
3.0 Towards a New Functional Anatomy of Language: The Present Issue
Richard Wise and Sophie Scott
MRC Cyclotron Unit, Hammersmith Hospital, London, and
Institute of Cognitive Neurosciences, University College London
Organisation of Cortical Systems for Speech Perception
Using a variety of speech, speech-like and non-speech stimuli, PET and fMRI studies have demonstrated that the lateral left superior temporal gyrus responds specifically to the phonetic content of speech or speech-like stimuli. PET studies have further demonstrated that the anterior part of the left superior temporal sulcus (STS) responds specifically to intelligible speech: this is a region readily overlooked in fMRI studies because of the presence of susceptibility artefact. PET and fMRI studies have also demonstrated auditory processing of speech in the left posterior temporal cortex. One part, directed along the supratemporal cortical plane, responds to both non-speech and speech sounds, including the sound of the speaker’s own voice. Activity in its most posterior and medial part, at the junction with the inferior parietal lobe, is linked to speech production rather than perception: this region co-activates with the left anterior insula and the most posterior part of the left inferior frontal gyrus during both propositional and non-propositional speech production. A more lateral and ventral region, in the posterior left superior temporal sulcus, responds both to an external source of speech and to the recall of lists of words during verbal fluency tasks. The results are compatible with a hypothesis that these adjacent regions of the posterior superior temporal cortex are specialised for processes involved in the mimicry of sounds, including repetition. The overall conclusion from these results is that there is a left-lateralised anterior stream of auditory processing for word recognition, and a posterior stream that interfaces with speech production, a system responsible, during language acquisition, for speech rehearsal and the laying down of long term lexical memories.
The perception of sentences and narratives, but not single words, activates the left temporo-parieto-occipital junction. We speculate that this area is responsible for ‘on-line’ verbal short term memory, necessary for relating the different, serially-perceived elements of sentences, words, phrases and clauses, to enable comprehension of sentences, particularly those with complex grammatical structures.
The lateral part of the right superior temporal cortex is also activated by the perception of speech. Using appropriate stimuli it is proving possible to demonstrate that it is the non-verbal components of communication, such as pitch and intonation, that are responsible for this response: this accords well with lesion data.
The basal language area, the left mid and anterior fusiform gyrus, which probably comprises the lateral parahippocampal and perirhinal cortex, is more strongly activated by high imageable than low imageable nouns. Activation of this area is most readily observed when subjects listen to narratives. This is multi-modal association cortex, and is, we speculate, a region involved in the semantic processing of speech and language, with some category-specificity (object > abstract words). It has strong reciprocal anatomical connections with the STS. The anterior STS also projects to the anterior temporal pole, and both infero-lateral and rostral prefrontal cortex. Co-activation of all these regions is observed when subjects perform explicit semantic tasks on verbal stimuli, when a choice has to signalled after mapping several possible semantic associations of a word with the subject’s interpretation of the choice he is being asked to make (e.g. can a human be a lemon?). In contrast, on-line perception of straightforward sentences and narratives activates only temporo-parietal regions, with no premotor or prefrontal cortical involvement.
Therefore, although incomplete, a picture is emerging of specific streams of processing during speech perception, and the onward projections from these streams can be partially segregated into those that interface with the premotor systems responsible for speech production and those that link with multi-modal association cortex responsible for accessing word meaning and holding verbal elements in short term memory. At the highest level, when an overt or covert decision must be made between of one of several meanings of an ambiguous verbal stimulus, amodal rostral prefrontal cortex is involved.
Dana Boatman
Johns Hopkins University
Neural Bases of Auditory Language Processing:
New Evidence from Lesion Studies
Outline
I. The traditional left hemisphere (unilateral) model of speech perception, first proposed in the 1960s, is still widely accepted today.
II. This model is based largely on studies of left hemisphere stroke patients with posterior cortical lesions and associated receptive aphasias.
III. Discuss limitations of extrapolating from stroke studies to neural bases of speech perception, including issues of bilateral subcortical auditory projections, lesion size, lesion location, presbycusis, and lack of premorbid data.
IV. Describe new generation of clinical lesion studies of speech perception that avoid limitations of traditional stroke models. Discuss particular lesion techniques, including cortical stimulation, Wada testing, hemispherectomy, transcranial magnetic stimulation.
V. Review recent speech perception lesion studies that challenge the traditional left hemisphere model of speech perception by showing that:
1) lesions of Wernicke's area do not necessarily produce speech perception (phonological) deficits, but instead may selectively impair access to lexical-semantic information from audition (discussion of cortical stimulation studies, re-analysis of stroke studies).
2) the speech perception capabilities of the right hemisphere have been underestimated (discuss recent Wada studies, aphasia recovery studies; hemispherectomy studies);
3) under certain listening conditions, speech perception appears to require both hemispheres (discuss two of our recent hemispherectomy studies; supporting evidence from neuroimaging literature and electrophysiology (PET and fMRI studies, Wise & Scott chapter) and how data provide support for claims by Hickok & Poeppel 2000).
VI. Argue for revision/rejection of traditional left hemisphere model of speech perception warrants based on recent lesion evidence. Revised model to include a critical role for both hemispheres in speech perception, recognition that different cortical circuits may be invoked for different speech perception tasks, and greater intra-hemispheric interaction between different cortical regions involved in speech perception.
Nina F. Dronkers1,2, David P. Wilkins3, Robert D. Van Valin Jr.4, Brenda B. Redfern1,2 and Jeri J. Jaeger4
1 VA Northern California Health Care System
2 University of California, Davis
3 Max-Planck-Institute for Psycholinguistics, Nijmegen
4 State University of New York at Buffalo
Lesion Analysis of Word- and Sentence-Level Comprehension Deficits in Aphasia
1.0 Introduction
2.0 Lexical- and Sentence-Level Comprehension Deficits in Aphasia: A Critical Review
3.0 Experiment 1: Neuroanatomy of Single Word Comprehension Deficits in Aphasia
4.0 Experiment 2: Neuroanatomy of Sentence Comprehension Deficits in Aphasia
5.0 Converging Evidence from Other Methodologies
6.0 Relation Between Comprehension and Production Systems
7.0 Summary and Conclusions
Michael Ullman
Georgetown University
The contribution of brain memory circuits to language
Our use of language depends upon two capacities: a mental lexicon of memorized words, and a mental grammar of rules that underlie the sequential and hierarchical composition of lexical forms into predictably structured larger words, phrases, and sentences.
The Declarative/Procedural model posits that the lexicon/grammar distinction in language is tied to the distinction between two well-studied brain memory systems. On this view, the memorization and use of at least simple words (those with non-compositional — that is, arbitrary — form-meaning pairings) depends upon an associative memory of distributed representations that is subserved by temporal-lobe circuits previously implicated in the learning and use of fact and event knowledge. This “declarative memory” system appears to be specialized for learning arbitrarily-related information (i.e., for associative binding). In contrast, the acquisition and use of aspects of symbol-manipulating grammatical are subserved by frontal/basal-ganglia circuits previously implicated in the implicit (non-conscious) learning and expression of motor and cognitive “skills” (e.g., from simple motor acts to skilled game playing). This “procedural” system may be specialized for computing sequences.
This novel view of lexicon and grammar offers an alternative to the two main competing theoretical frameworks. It shares the perspective of traditional Dual-Mechanism theories in positing that the mental lexicon and a symbol-manipulating mental grammar are subserved by distinct computational components that may be linked to distinct brain structures. However, it diverges from these theories where they assume components dedicated to each of the two language capacities (that is, domain-specific), and in their common assumption that lexical memory is a rote list of items. Conversely, while it shares with Single-Mechanism theories the perspective that the two capacities are subserved by domain-independent computational mechanisms, it diverges from them where they link both capacities to a single associative memory system with broad anatomic distribution.
The Declarative/Procedural model, but neither traditional Dual-Mechanism nor Single-Mechanism models, predicts double dissociations between lexicon and grammar, with associations among associative memory properties, memorized words and facts, and temporal-lobe structures, and among symbol-manipulation properties, grammatical rule-products, motor skills, and frontal/basal-ganglia structures.
Although data from studies of syntactic computation (in particular, of syntactic structure building), and of phonological sequencing are presented, the paper focuses on morphology. The investigation of morphologically complex forms allow one to contrast lexicon and grammar while holding other factors relatively constant. Morphological transformations that are (largely) unproductive (e.g., in go-went, solemn-solemnity) are hypothesized to be memorized in declarative memory. These have been contrasted with morphological transformations that are fully productive (e.g., in walk-walked, happy-happiness), whose computation is posited to be dependent upon grammatical rules subserved by the procedural system.
Multiple lines of evidence are presented, from a number of languages (including English, Italian, German, and Japanese), using a range of psycholinguistic and neurolinguistic approaches, including: the examination of frequency, neighborhood, imageability, working memory, and priming effects; the behavioral testing of patients with developmental, acquired, and neurodegenerative disorders, including Specific Language Impairment, Phenylketonuria, Williams syndrome, posterior and anterior aphasia, and Alzheimer’s, Parkinson’s, and Huntington’s diseases; and neuroimaging, using fMRI, EEG/ERPs, and MEG. In addition, evidence will be presented that suggests differences in the relative lexical/grammatical reliance on the two memory systems between men and women, and between first and second language speakers. Finally, implications from the model for therapeutic approaches to language disorders are discussed, and supporting evidence is presented.
Angela Friederici
Max Planck Institute for Cognitive Neuroscience
Neural Basis of Syntactic Processing
Traditionally left frontal and temporal language areas have been functionally defined as supporting production and comprehension, respectively. More recently the distinction has been made between syntactic and lexical-semantic processes. Here the view will be put forward that both frontal and temporal brain areas are involved in syntactic processes during language comprehension. The functional specification of these areas describes the anterior part of the superior temporal gyrus (planum polare) and the inferior frontal gyrus to support syntactic processing during sentence comprehension as indicated by fMRI.
Both areas are involved in early structure building processes during comprehension as shown by MEG results. The particular function the planum polare might play, is to access word category information which is a necessary pre-condition to engage structure building processes. This information provides immediate access to all of the syntactic rules involving a given category and is thus, the basis for rule application procedures possibly supported by the inferior frontal gyrus. The high automaticity of these procedures only leads to minimal activation in the frontal region when processing simple sentences.
The inferior frontal gyrus (BA 44) is shown to come into play more strongly when syntactic processes become more demanding, be it due to an increase in syntactic memory requirement as in long distance dependencies or be it due to an increase in the attentional demands on syntactic features as in certain tasks.
Willem J.M. Levelt and Peter Indefrey
Max Planck Institute for Psycholinguistics
The spatial and temporal signatures of language production components
In a recent meta-analysis (Indefrey & Levelt, 2000) we have identified a left-lateralized cerebral network for the core processes of word production. The brain activation patterns that were reliably induced by different word production tasks turned out to be compatible with a theory-driven analysis of the functional processing components of the respective tasks. This allowed a tentative identification of the neural correlates of single processing components, such as lexical word-form retrieval and post-lexical phonological encoding.
In the present work we will investigate the implications of these findings in two directions. (i) Relations of cerebral regions to specific processing components predict not only their activation in a certain set of experimental tasks, but also time-windows in which this activation should be observed. We will examine whether the available electrophysiological and magneto-encephalographic data on language production are compatible with the predicted time-windows. (ii) There is solid evidence from psycholinguistic experiments demonstrating an influence of language comprehension processes on simultaneously occurring language production processes and vice versa. There is also evidence, however, that the two processing streams do not simply recruit all processing components in the opposite direction. It is a controversial and to date unresolved question, which processing steps are bidirectional, involving shared processing components, and which are not. In the present work, we will extend the meta-analysis procedure developed in Indefrey & Levelt (2000) to investigate the locations of cerebral activations during selected language comprehension and production tasks in a common anatomical reference space. This will allow the identification of sets of cerebral areas that are shared between different production and comprehension tasks. In a further step we will attempt to link the functional neuroanatomical data to the shared processing components to be identified in a theory-based task analysis.
1.0 The functional organization of word production
1.1. The components of word production
1.2. The time course of word production
1.3. Effects of comprehension on word production
2.0 Cerebral regions involved in word production and comprehension - a meta-analysis
2.1. A component analysis of word production and comprehension tasks
2.2. Meta-analysis procedures
2.3. Neural correlates of task-specific lead-in processes
2.4. Neural correlates of the core processes of word production
2.5. Shared neural correlates of word production and comprehension processes
3.0 The time course of cerebral activations during word production
3.1. Time-windows of interaction between word production and comprehension
3.2. Integrating temporal and spatial information - a tentative flow-chart of word production
H. Damasio, A. Damasio, D. Tranel, T. Grabowski and R. Adolphs
University of Iowa
Neural Correlates of Category Related Naming
1.0 Introduction
2.0 Theoretical framework
3.0 Methods issues
4.0 Evidence from lesion studies
5.0 Evidence from functional imaging
6.0 A neuroanatomical architecture for word retrieval
Gregory Hickok and David Poeppel
University of California, Irvine, and University of Maryland, College Park
Dorsal and Ventral Streams in Speech and Language Processes
1.0 Preliminaries
1.1 A Perspective on Linguistic Specificity in Neural Organization
1.2 On Tasks and Goals in Language Processing and Brain Mapping
2.0 Overview of the Model
3.0 Task Dissociations in "Speech Perception"
3.1 Aphasia
3.2 Split Brain
3.3 Neuroimaging
3.4 Illiteracy
4.0 The Ventral Stream(s?)
4.1 Bilateral organization at the lowest levels of the processing hierarchy
4.1.2 Wernicke's footnote
4.1.2 Brief review of evidence: aphasia, word deafness, split brains, imaging, stimulation
4.2 Asymmetries embedded in bilateral organization
4.3 The sound-meaning interface
4.4 Speculations on grammatical processes in anterior superior temporal lobe
5.0 The Dorsal Stream
5.1 Sensory-motor integration in the parietal lobe
5.2 Sensory-motor integration in speech
5.2.1 Developmental requirements
5.2.2 Operation in the adult
5.2.2.1 Repetition of pseudowords
5.2.2.2 Altered auditory feedback experiments
5.2.2.3 Shadowing latency suggests efficient S-M system
5.2.3 Role in phonological working memory
5.2.4 Role in sub-lexical speech tasks
5.3 Localization
5.3.1 Working memory tasks
5.3.2 Sub-lexical speech tasks
6.0 Perception-Production Overlap in Posterior "Sensory" Cortex
6.1 Evidence from aphasia
6.2 Evidence from stimulation mapping
6.3 Evidence from neuroimaging
6.4 Evidence from cytoarchitectonics
6.5 Relation to Motor-Theory of Speech Perception
6.6 On mirror neurons
7.0 Understanding Aphasia
7.1 Word deafness
7.2 Conduction aphasia
7.3 Transcortical sensory aphasia
7.4 Wernicke's aphasia
References
Aboitiz, F., & García V., R. (1997). The evolutionary origin of language areas in the human brain. A neuroanatomical perspective. Brain Research Reviews, 25, 381-396.
Anderson, J.M., Gilmore, R., Roper, S., Crosson, B., Bauer, R.M., Nadeau, S., Beversdorf, D.Q., Cibula, J., Rogish III, M., Kortencamp, S., Hughes, J.D., Gonzalez Rothi, L.J., & Heilman, K.M. (1999). Conduction aphasia and the arcuate fasciculus: A reexamination of the Wernicke-Geschwind model. Brain and Language, 70, 1-12.
Blumstein, S. (1995). The neurobiology of the sound structure of language. In M.S. Gazzaniga (Ed.), The cognitive neurosciences (pp. 913-929). Cambridge, MA: MIT Press.
Boatman, D., Hart, J.J., Lesser, R.P., Honeycutt, N., Anderson, N.B., Miglioretti, D., & Gordon, B. (1998). Right hemisphere speech perception revealed by amobarbital injection and electrical interference. Neurology, 51, 458-464.
Damasio, H., Grabowski, T.J., Tranel, D., Hichwa, R.D., & Damasio, A.R. (1996). A neural basis for lexical retrieval. Nature, 380, 499-505.
Dronkers, N.F. (1996). A new brain region for coordinating speech articulation. Nature, 384, 159-161.
Dronkers, N.F., Redfern, B.B., & Knight, R.T. (2000). The neural architecture of language disorders. In M.S. Gazzaniga (Ed.), The new cognitive neurosciences (pp. 949-958). Cambridge, MA: MIT Press.
Hickok, G., & Poeppel, D. (2000). Towards a functional neuroanatomy of speech perception. Trends in Cognitive Sciences, 4, 131-138.
Indefrey, P., & Levelt, W.J.M. (2000). The neural correlates of language production. In M.S. Gazzaniga (Ed.), The new cognitive neurosciences (pp. 845-865). Cambridge, MA: MIT Press.
Norris, D., & Wise, R. (2000). The study of prelexical and lexical processes in comprehension: Psycholinguistics and functional neuroimaging. In M.S. Gazzaniga (Ed.), The new cognitive neurosciences (pp. 867-880). Cambridge, MA: MIT Press.
Ullman, M.T., Corkin, S., Coppola, M., Hickok, G., Growdon, J.H., Koroshetz, W.J., & Pinker, S. (1997). A neural dissociation within language: Evidence that the mental dictionary is part of declarative memory, and that grammatical rules are processed by the procedure system. Journal of Cognitive Neuroscience, 9, 266-276.