An fMRI investigation of syllable sequence production
Introduction
Fluent speech requires a robust serial ordering mechanism to combine a finite set of discrete learned phonological units (such as phonemes or syllables) into larger meaningful expressions of words and sentences. Lashley (1951) posed the problem of serial order in behavior, asking how the brain organizes and executes smooth, temporally integrated behaviors such as speech and rhythmic motor control. His proposal for the “priming of expressive units,” or parallel, co-temporal activation of the items in a behavioral sequence prior to execution, has been supported in studies of speech production by bountiful data related to linguistic performance errors (e.g., MacKay, 1970, Fromkin, 1980, Gordon and Meyer, 1987), by reaction time experiments (e.g., Klapp, 2003), and by the demonstration of anticipatory and perseveratory co-articulation (e.g., Ohman, 1966; Hardcastle and Hewlett, 1999).
The problem of serial order in speech production can be considered at multiple levels. Phonemes, for example, might be manipulated to form syllables and words, where each phonemic token is learned and stored with corresponding auditory and/or orosensory consequences (see, for example, the DIVA model of speech production; Guenther, 1995, Guenther et al., 2006, which provides a computational account for how such tokens can be learned and produced). Data also suggest that syllable or word-sized tokens can be learned such that they may be efficiently executed as single motor chunks, forming a mental syllabary (Levelt and Wheeldon, 1994, Levelt et al., 1999, Cholin et al., 2006); these larger chunks might then serve as manipulable tokens for speech sequence planning.
In addition to organizing sequences of planned sounds within a memory buffer, speech production requires a mechanism to initiate or release items to the motor apparatus at precise times. Speakers can typically produce up to six to nine syllables (20 to 30 segments) per second, which is faster than any other form of discrete motor behavior (Kent, 2000). A system that coordinates the timed release of each discrete item in the planned sequence of speech is, therefore, of critical importance to fluent performance.
While the formulation of spoken language plans has been widely studied at a conceptual level (see, e.g., Levelt, 1989; Levelt et al., 1999), relatively little is known about the neural representations of those plans or about the cortical and subcortical machinery that guides the serial production of speech. Clinical studies have suggested that damage to the anterior insula or neighboring inferior frontal areas (Dronkers, 1996, Hillis et al., 2005, Tanji et al., 2001), supplementary motor area (Jonas, 1981, Jonas, 1987, Ziegler et al., 1997, Pai, 1999), basal ganglia (Pickett et al., 1998, Ho et al., 1998), or cerebellum (Riva, 1998, Silveri et al., 1998) may lead to deficits in sequencing and/or initiation of speech plans. Such deficits appear in various aphasias and apraxia of speech (AOS). Literal or phonemic paraphasias, in which “well-formed sounds or syllables are substituted or transposed in an otherwise recognizable target word” (Goodglass, 1993), exist in many aphasic patients including Broca's and (most commonly) conduction aphasics. AOS, a speech-motor condition1, has been attributed to damage to the left precentral gyrus of the insula (Dronkers, 1996), as well as the inferior frontal gyrus, subcortical structures, or posterior temporal/parietal regions (Hillis et al., 2005, Peach and Tonkovich, 2004, Duffy, 1995). Ziegler (2002) presents an excellent review of theoretical models of AOS.
Only a small portion of the large functional neuroimaging literature related to speech and language has dealt with overt speech production. Within that body, very few studies have explicitly addressed sequencing demands during overt speech. Riecker et al. (2000b) examined brain activations evoked by repetitive production of stimuli of varying complexity: consonant vowel syllables (CVs), CCCVs, CVCVCV non-word sequences, and CVCVCV words. This study found that production of none of the stimulus types (compared to a resting baseline condition) resulted in significant activations in the SMA or insula; activation was instead largely restricted to the primary sensorimotor areas. Only CCCV production led to significant activation of the cerebellum. Production of the multi-syllabic items led to a more limited and lateralized expanse of activation in the banks of the central sulcus than did production of single syllables.
Shuster and Lemieux (2005) compared production (both overt and covert) of multi-syllabic and mono-syllabic words following the presentation of an auditory exemplar. For overt speech, additional activation was found in the left inferior parietal lobe, inferior frontal gyrus, and precentral gyrus for multi-syllabic versus mono-syllabic words. Mono-syllabic words resulted in greater activation of the left middle frontal gyrus (BA46). The results for covert speech were somewhat dissimilar; for example, in covert speech, there was greater activation of the left middle frontal gyrus for multi-syllable words and greater activation in the left precentral gyrus for mono-syllable words. A consistent finding was that multi-syllable words caused additional activation in left inferior parietal areas (BA40), a region the authors suggest to be involved in speech programming. In comparing the results of this study to that of Riecker et al. (2000b), it is difficult to develop a consistent account for the effects of sequential complexity on the speech production system.
In the present experiment, we sought to clarify how the speech system organizes and produces sequences of speech sounds. While the DIVA model of speech production makes predictions about brain activations in the executive speech motor system (Guenther et al., 2006, Guenther, in press), it does not address brain regions likely to be responsible for sequence planning. Based on clinical observations and studies of other non-speech sequential motor control tasks, we expected to observe additional responses to additional stimulus complexity in a network of brain regions outside of the primary sensorimotor areas (and other regions treated by the DIVA model), including the prefrontal cortex, basal ganglia, anterior insula, supplementary motor area, and cerebellum. Blood oxygenation level-dependent (BOLD) functional magnetic resonance imaging (fMRI; see Ogawa et al., 1990, Belliveau et al., 1991, Kwong et al., 1992) was used to measure responses to speech sequences of varying complexity at both the sub- and suprasyllabic levels and in both preparatory and overt speech production tasks. We employed an “event-triggered” design with GO and NOGO trials that offered many benefits over previous methods (see Discussion). We discuss the results in terms of the necessary mechanisms for sequencing and initiation in fluent speech production.
Section snippets
Subjects
Thirteen right-handed native English speakers (ages 22–50 years, mean 28.7 years, six females) with no history of neurological, speech, language, or hearing disorders participated. Written informed consent was obtained according to the Boston University Institutional Review Board and the Massachusetts General Hospital Human Research Committee.
Experimental protocol
Tasks consisted of preparing to produce (NOGO trials) and overtly producing (GO trials) three syllable sequences. The linguistic content of the stimuli
Acoustic analysis
The mean acoustic duration and between-subject standard deviation (in ms) for utterances of each stimulus type were as follows: S_seq, S_syl: 993 ± 215; C_seq, S_syl: 1006 ± 186; S_seq, C_syl: 1195 ± 209; C_seq, C_syl: 1332 ± 155. The difference between S_seq, S_syl and C_seq, S_syl was not significant. All other pair-wise differences were significant (P < 0.05).
Basic speech production network
Production of each of the stimulus types was individually contrasted with the baseline condition. Group results showed regions of
Discussion
In this study, we sought to better understand the neural substrates for planning and producing sequences of simple speech sounds, a faculty that is ubiquitous in normal discourse. This topic has received relatively little attention in the neuroimaging literature to date, with most studies of language production focusing on aspects of word generation and production (reviewed in Indefrey and Levelt, 2000, Turkeltaub et al., 2002) or on other aspects of verbal output such as speaking rate (
Acknowledgments
This research was supported by the National Institute on Deafness and other Communication Disorders (R01 DC02852, F. Guenther PI). Imaging was performed at the Athinoula A. Martinos Center for Biomedical Imaging; this work was made possible by grants from the National Center for Research Resources (P41RR14075) and the MIND institute. The authors would like to thank Daniel Bullock, Satrajit Ghosh, Jason Tourville, Alfonso Nieto-Castanon, Julie Goodman, and Larry Wald for their assistance with
References (152)
- et al.
The contribution of the insula to motor aspects of speech production: a review and a hypothesis
Brain Lang.
(2004) - et al.
Functional architecture of basal ganglia circuits: neural substrates of parallel processing
Trends Neurosci.
(1990) Circuitry and functional aspects of the insular lobe in primates including humans
Brain Res. Rev.
(1996)- et al.
Overt verbal responding during fMRI scanning: empirical investigations of problems and potential solutions
NeuroImage
(1999) - et al.
Experimental designs and processing strategies for fMRI studies involving overt verbal responses
NeuroImage
(2004) - et al.
How laminar frontal cortex and basal ganglia circuits interact to control planned and reactive saccades
Neural Netw.
(2004) - et al.
Phonetic perceptual identification by native and second-language speakers differentially activates brain regions involved with acoustic phonetic processing and those involved with articulatory-auditory/orosensory internal models
NeuroImage
(2004) - et al.
Cerebrocerebellar networks during articulatory rehearsal and verbal working memory tasks
NeuroImage
(2005) - et al.
Effects of syllable frequency in speech production
Cognition
(2006) - et al.
Modality effects in verbal working memory: differential prefrontal and parietal responses to auditory and visual stimuli
NeuroImage
(2004)
Cortical surface-based analysis: I. Segmentation and surface reconstruction
NeuroImage
Functional MRI studies of spatial and nonspatial working memory
Cogn. Brain Res.
Cerebral regions associated with verbal response initiation, suppression and strategy use
Neuropsychologia
Over de snelheid van psychische proessen (On the speed of mental processes)
Acta Psychol.
Neuroimaging studies of speech: an overview of techniques and methodological approaches
J. Commun. Disord.
Cortical surface-based analysis: II. Inflation, flattening, and a surface-based coordinate system
NeuroImage
Conjunction revisited
NeuroImage
Thresholding of statistical maps in functional neuroimaging using the false discovery rate
NeuroImage
Imaging speech production using fMRI
NeuroImage
Neural modeling and imaging of the cortical interactions underlying syllable production
Brain Lang.
Serial position effects in nonword repetition
J. Mem. Lang.
Mutism and auditory agnosia due to bilateral insular damage — role of the insula in human communication
Neuropsychologia
Combining voxel intensity and cluster extent with permutation test framework
NeuroImage
Recoding, storage, rehearsal and grouping in verbal short-term memory: an fMRI study
Neuropsychologia
Sequence heterogeneity in parkinsonian speech
Brain Lang.
The supplementary motor region and speech emission
J. Commun. Disord.
The efferent and efferent connections of the supplementary motor area
Brain Res.
Prefrontal cortex guides context-appropriate responding during language production
Neuron
Load- and practice-dependent increases in cerebro-cerebellar activation in verbal working memory: an fMRI study
NeuroImage
Selection of actions in the basal ganglia-thalamocortical circuits: review and model
Int. J. Psychophysiol.
Cognitive and language functions of the human cerebellum
Trends Neurosci.
Do speakers have access to a mental syllabary?
Cognition
Anticipatory activity in primary motor cortex codes memorized movement sequences
Neuron
Spoonerisms: the structure of errors in the serial order of speech
Neuropsychologia
Basal ganglia and cerebellar loops: motor and cognitive circuits
Brain Res. Rev.
The basal ganglia: focused selection and inhibition of competing motor programs
Prog. Neurobiol.
Basal ganglia intrinsic circuits and their role in behavior
Curr. Opin. Neurobiol.
Functional imaging during speech production
Acta Psychol.
Functional magnetic resonance imaging of verbal fluency and confrontation naming using compressed image acquisition to permit overt responses
Hum. Brain Mapp.
Speech rate and rhythm in cerebellar dysarthria: an acoustic analysis of syllabic timing
Folia Phoniatr. Logop.
Speech deficits in ischaemic cerebellar lesions
J. Neurol.
Temporal organization of ”internal speech” as a basis for cerebellar modulation of cognitive functions
Behav. Cogn. Neurosci. Rev.
Parallel organization of functionally segregated circuits linking basal ganglia and cortex
Annu. Rev. Neurosci.
Parallel processing of serial movements in prefrontal cortex
Proc. Natl. Acad. Sci.
Neural activity in prefrontal cortex during copying geometrical shapes: I. Single cells encode shape, sequence, and metric parameters
Exp. Brain Res.
Dissociation of storage and rehearsal in verbal working memory
Psychol. Sci.
Working Memory
Functional mapping of the human visual cortex by magnetic resonance imaging
Science
Controlling the false discovery rate: a practical and powerful approach to multiple testing
J. R. Stat. Soc., Ser. B Methodol.
Magnetic field changes in the human brain due to swallowing or speaking
Magn. Reson. Med.
Cited by (403)
Microstructural white matter changes underlying speech deficits in Parkinson's disease
2024, Brain and LanguageRapid neural changes during novel speech-sound learning: An fMRI and DTI study
2023, Brain and LanguageNeurocomputational modeling of speech motor development
2023, Journal of Child Language