About

Gregory Hickok is a professor involved in the Auditory & Language Neuroscience Lab at the University of California, Irvine. His research focuses on the functional anatomy of language, the neural organization of speech production and perception, and the neural organization of gestural communication, including evidence from sign language. His work encompasses the neural basis of language and music perception, as well as computational models of aphasia and the neural mechanisms underlying speech and gesture. Hickok's contributions include exploring the neural organization of speech and language, utilizing techniques such as electrocorticography (ECoG), and investigating the neural substrates involved in language processing and communication.

Research topics

• Computer Science
• Psychology
• Linguistics
• Medicine
• Natural Language Processing
• Cognitive psychology
• Cognitive science
• Artificial Intelligence
• Audiology
• Neuroscience
• Speech recognition
• Programming language
• Developmental psychology

Selected publications

• 1296 Shared Cortical Language Networks With Convergent Hierarchical Network Dynamics for Lexicosemantic Processing in Comprehension and Naming

  Neurosurgery · 2025-03-14

  article

  INTRODUCTION: Theoretical models suggest that spoken and written language engage a shared lexicosemantic processing network in perception and production, yet convergent neural mechanisms are unclear. METHODS: 65 ECoG patients completed auditory (AN) and orthographic (ON) naming. We analyzed gamma activity (70-115Hz) with mixed-effects multilevel analyses to identify the lexicosemantic processing network during comprehension and naming. We mapped network dynamics using autoregressive hidden Markov models (ARHMM). We used direct cortical stimulation (DCS) to attribute causality to critical nodes. RESULTS: At speech onset, activation of superior temporal gyrus (pSTG) was followed by superior temporal sulcus (pSTS) and middle temporal gyrus (pMTG). For each written word, visual cortex activity was followed by activation of lexical (fusiform gyrus, Fus; pSTS; pMTG) and phonological (intraparietal sulcus, IPS; pSTG) reading routes. Both modalities engaged posterolateral temporal cortex (pLTC) for comprehension, and activity was correlated with phrasal composition (p<0.01) implicating it in compositional semantics. The last word activated a shared network (pLTC; Fus; IPS; pars triangularis, pTr) for naming. ARHMM isolated 5 states for AN and 6 for ON with 3 convergent states. The first convergent state occurring at stimulus offset was characterized by outflow from pLTC, Fus, IPS, and pTr, and state duration was correlated with reaction time (p<0.001) implicating it in lexical access. Lastly, during stimulus presentation, DCS of Heschl’s gyrus disrupted listening, while DCS of planum temporale and pLTC disrupted listening and reading. At stimulus offset, DCS of pLTC, Fus, IPS, and pTr disrupted AN and ON. CONCLUSIONS: Juxtaposing network dynamics of multimodal lexicosemantic processing in speech perception and production informs our understanding of specialized and shared language networks providing new insights to facilitate designs of neural prosthetics for language disorders.

  PublisherDOI

• A review of sentence elicitation techniques for assessing grammatical deficits in aphasia

  Aphasiology · 2025-10-29

  article
  PublisherDOI

• Distinct patterns of syntactic errors doubly dissociate in chronic post-stroke aphasia

  2025-07-15

  preprintOpen access

  Background: The lesion correlates of syntactic deficits in aphasia remain poorly understood. Previous studies have suggested that distinct error types in expressive syntax, such as paragrammatic and agrammatic speech, may be associated with damage to different brain regions, but the specific lesion correlates of these errors have not been fully delineated.Objective: To identify the lesion correlates associated with distinct expressive syntactic error types in individuals with chronic post-stroke aphasia, using a novel utterance-level analysis of spontaneous speech errors.Methods: We analyzed spontaneous speech samples from individuals with chronic aphasia, categorizing errors into hierarchical (paragrammatic-like) and linearization (agrammatic-like) errors. Lesion-symptom mapping was conducted to identify brain regions associated with these error types. The analysis was based on a two-stage model of sentence production, with hierarchical processing and linearization analyzed as distinct stages.Results: Lesion clusters in the medial superior temporal sulcus and inferior parietal lobe were associated with hierarchical errors, while large frontal lesions, including those in the inferior and middle frontal lobe, were associated with linearization errors. These results provide support for a two-stage model of syntactic encoding, with distinct neural correlates for hierarchical and linearization processes. Conclusion: Our findings suggest a dichotomy of the lesion basis for syntactic deficits in aphasia, suggesting that distinct brain regions contribute to different stages of syntactic encoding. Hierarchical processing seems to be supported by posterior temporal-parietal regions, while linearization seems to be supported by frontal regions. These results suggest that the diagnosis and treatment of aphasia should take into account the existence of distinct syntactic production deficits resulting from different patterns of brain damage.

  PublisherOA PDFDOI

• ‘Wired for Words: The Neural Architecture of Language,’ an excerpt

  The Transmitter · 2025-01-01

  article1st authorCorresponding
  PublisherDOI

• Mapping of critical prosodic and phonetic networks in post-stroke apraxia of speech

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-12-19

  articleOpen accessSenior author

  Purpose: Many have made proposals to better diagnose and/or classify post-stroke apraxia of speech (AOS), with some arguing for the separation of AOS into behavioral subtypes. Recent studies of primary progressive AOS have promoted a separation of prosodic and phonetic subtypes, aligning with a dual-motor coordination model separating the neural substrates of prosodic and phonetic function. Motivated by the limited corroboration of these subtypes in post-stroke AOS, here we present mapping results in a cohort of stroke survivors aiming to identify distinct neural substrates for prosodic and phonetic aspects of speech motor coordination. Methods: = 127; 64 with AOS) received speech-language evaluation and neuroimaging at the Center for the Study and Treatment of Aphasia Recovery (C-STAR). AOS severity was quantified via the Apraxia of Speech Rating Scale (ASRS). We utilized a novel lesion-symptom mapping technique with an emphasis on prediction that identifies ensembles of regions supporting performance in the prosodic and phonetic domains. Results: An ensemble of networks supporting prosodic function localized to dorsal and ventral (but primarily dorsal) sensorimotor cortex, as well as a distributed network of white matter pathways connecting Rolandic cortex to auditory regions and cerebellum, emphasizing the role of auditory feedback processing and laryngeal control in supporting prosodic function. A separate but partially overlapping network supporting phonetic function localized primarily to ventral Rolandic cortex and the arcuate fasciculus. Conclusions: This work represents the first mapping of prosodic and phonetic subtypes in post-stroke AOS in a large cohort of individuals. We hope our results motivate the development of assessment and treatment techniques individually targeting prosodic and phonetic functioning to better serve individuals with AOS and facilitate clinical discussion of the disorder.

  PublisherDOI

• Take a Good Hard Look at Your Mirror Neurons

  2025-01-30

  articleOpen access1st authorCorresponding
  PublisherOA PDFDOI

• Acute temporal lesions are associated with phonological word verification errors

  Brain Network Disorders · 2025-02-28 · 1 citations

  articleOpen access

  Within the dual stream model for language, the bilateral dorsal superior temporal gyrus (STG) is associated with spectro-temporal analysis while the mid-post superior temporal sulcus is associated with processing of higher-level phonological codes. However, the true lateralization of functions needed for phonological discrimination at the word level remains unsettled. The aim of the present work was to determine if individuals with acute stroke primarily involving the left STG and middle temporal gyrus (MTG) demonstrated poorer discrimination ( d’ ) between phonologically related words in a word-picture verification task than those with left hemisphere lesions outside the temporal lobe and compared to those with right hemisphere stroke. The latter would support a left bias and provide an estimate of the magnitude. One hundred fourteen individuals with acute stroke completed both acute magnetic resonance imaging and a word-picture verification task with phonological and semantic foils. Eighty-nine participants had left hemisphere stroke (51 included the temporal lobe). Twenty participants had right hemisphere stroke (14 included the temporal lobe). Five participants with bilateral stroke were included (3 included the bilateral temporal lobes). Quantile regression was performed controlling for overall lesion volume, age, and sex. While more than half of patients performed the task well, patients with left temporal lobe lesions were more likely to perform poorly. These results confirm a role for the left superior temporal lobe, including the STG, in phonological processing during word recognition, and indicate that a subset of individuals (<50%) have varying degrees of left bias in discrimination of phonologically similar words in the STG.

  PublisherDOI

• More Reflections on Mirror Neurons

  2025-02-04

  articleOpen access1st authorCorresponding

  The authors trained non-musicians to play a piece of music on a keyboard.They then scanned the subjects using fMRI while they passively listened to the piece they learned or to a piece the hadn't learned to play.Both

  PublisherOA PDFDOI

• Audiovisual Synchrony in Left-hemisphere Brain-lesioned Individuals with Aphasia

  Journal of Cognitive Neuroscience · 2025-01-01

  articleSenior author

  We investigated the ability of 40 left-hemisphere brain-lesioned individuals with various diagnoses of aphasia to temporally synchronize the audio of a spoken word to its congruent video using a maximum-likelihood adaptive psychophysical procedure. We found a statistically significant effect of aphasia type, not explained by lesion volume, on measures of audiovisual (AV) synchrony. Brain-lesioned individuals with no symptoms of aphasia, and those with conduction aphasia performed on the synchrony task more similarly to age-matched neurotypical controls, whereas those with anomic aphasia performed the poorest. In addition, we examined the correlation between this ability and AV integration (fusion) and observed a significant correlation between measures of AV synchrony and fusion. An ROI analysis of stroke lesion maps showed that damage to the left posterior temporal regions adversely affected AV processing, although whole-brain univariate lesion-symptom mapping analyses did not yield any significant results. These findings contribute to a better understanding of the functional relationship between different AV processes in multimodal integration and their underlying cortical networks in the human brain.

  PublisherDOI

• Transfer learning via distributed brain recordings enables reliable speech decoding

  Nature Communications · 2025-10-01 · 2 citations

  articleOpen access

  Speech brain-computer interfaces (BCIs) combine neural recordings with large language models to achieve real-time intelligible speech. However, these decoders rely on dense, intact cortical coverage and are challenging to scale across individuals with heterogeneous brain organization. To derive scalable transfer learning strategies for neural speech decoding, we used minimally invasive stereo-electroencephalography recordings in a large cohort performing a demanding speech motor task. A sequence-to-sequence model enabled decoding of variable-length phonemic sequences prior to and during articulation. This enabled development of a cross-subject transfer learning framework to isolate shared latent manifolds while enabling individual model initialization. The group-derived decoder significantly outperformed models trained on individual data alone, enabling decoding robustness despite variable coverage and activation. These results highlight a pathway toward generalizable neural prostheses for speech and language disorders by leveraging large-scale intracranial datasets with distributed spatial sampling and shared task demands. Speech brain-computer interfaces face challenges scaling across individuals with different brain organization. Using minimally invasive recordings from 25 patients, the authors developed transfer learning methods that enable robust speech decoding even with incomplete brain coverage.

  PublisherOA PDFDOI

Recent grants

• Center for the Study of Aphasia Recovery (C-STAR)

  NIH · $48.5M · 2016–2027

• Neurobiology of Auditory Language Perception

  NIH · $7.2M · 1999–2016

• Integrative Functions of the Planum Temporale

  NIH · $5.1M · 2008–2022

Frequent coauthors

• Julius Fridriksson

  University of South Carolina

  60 shared

• Argye E. Hillis

  Johns Hopkins University

  55 shared

• Leonardo Bonilha

  Medical University of South Carolina

  36 shared

• Alexandra Basilakos

  University of South Carolina

  34 shared

• Kourosh Saberi

  32 shared

• Chris Rorden

  University of South Carolina

  31 shared

• William Matchin

  University of South Carolina

  28 shared

• Ursula Bellugi

  26 shared

Labs

• Auditory & Language Neuroscience LabPI