Summary: Neuroscientists have located the brain’s auditory lexicon—the neural catalog for spoken words—not where long-standing models predicted. Instead of being positioned behind primary auditory cortex, evidence now points to the front of primary auditory cortex, in a region called the Auditory Word Form Area (AWFA). This revision of cortical organization has important implications for understanding speech comprehension and for designing rehabilitation after brain injury.
A team at Georgetown University Medical Center used a sensitive imaging method to examine spoken-word processing and new word learning in 26 volunteers. Their results challenge a century-old assumption about where spoken-word representations reside and suggest fresh directions for research and therapy targeted at speech comprehension deficits, including those arising after stroke.
Key Facts:
- The auditory lexicon—neural representations that encode whole spoken words—was identified in the anterior portion of the primary auditory cortex, contrary to historical models that placed it posteriorly.
- Researchers applied functional MRI rapid adaptation (fMRI-RA), a technique that offers greater sensitivity than standard fMRI for detecting how the brain represents and learns auditory words.
- These findings refine our understanding of speech processing and could guide rehabilitation strategies for patients with language comprehension deficits following stroke or other brain injuries.
Source: Georgetown University Medical Center
New anatomical insight on spoken word representation
For more than a century, neuroscientific models placed the core process of spoken-word recognition behind the primary auditory cortex. That picture did not always align with clinical observations: many patients with speech recognition difficulties—especially stroke survivors—show patterns of impairment that were hard to reconcile with the posterior-location hypothesis. The Georgetown study, led by Maximilian Riesenhuber, PhD, now provides evidence that the auditory lexicon is situated more anteriorly, within the left anterior superior temporal gyrus in a region the authors identify as the Auditory Word Form Area (AWFA).
This discovery mirrors earlier work from the same laboratory showing a visual lexicon for written words in the Visual Word Form Area (VWFA) at the base of the left hemisphere. In both cases, familiar whole-word representations appear to be stored in modality-specific ventral stream regions, suggesting convergent strategies for representing words across vision and audition.
The study involved 26 participants who completed three rounds of fMRI scanning while performing spoken-word processing tasks. The investigators used functional MRI rapid adaptation (fMRI-RA), which detects fine-grained neural selectivity and the sharpening of representations that occurs with learning. Using fMRI-RA, they demonstrated that the AWFA contains neural populations highly selective for individual spoken words and that exposure to new auditory words sharpens these representations, indicating a mechanism for learning and familiarization.
Riesenhuber notes that the anterior location of the auditory lexicon provides a clearer anatomical target for future studies and interventions aimed at improving speech comprehension. “Since the early 1900s, scientists believed spoken word recognition occurred behind the primary auditory cortex, but that model did not align with many observations from patients with speech recognition deficits,” he says. “Finding an auditory lexicon toward the front of the brain gives us a new region to study and potentially target in rehabilitation.”
The team plans to explore how direct interventions at the AWFA might influence recovery from different types of brain injury, and to investigate how auditory and visual word systems interact. They are also applying similar methods to search for auditory lexica in brain systems involved in speech production.
Co-author Josef Rauschecker, PhD, DSc, highlights remaining open questions about how the brain monitors and refines spoken output. Auditory feedback helps speakers—both children learning their first language and adults acquiring additional languages—adjust pronunciation and accuracy, but the neural mechanisms that compare intended and produced words are not yet fully understood.
Authors on the paper include Maximilian Riesenhuber, Srikanth Damera, Josef Rauschecker, Lillian Chang, Plamen Nikolov, James Mattei, Suneel Banerjee, Patrick H. Cox, and Xiong Jiang from Georgetown University, with Laurie S. Glezer from San Diego State University. The study appears in Neurobiology of Language (published July 5, 2023).
Funding: This research was supported by National Science Foundation grant BCS-1756313 and NIH grant 1S10OD023561. Computational resources were provided through the Extreme Science and Engineering Discovery Environment (XSEDE), supported by NSF grant ACI-1548562.
About this linguistics and neuroscience research news
Author: Karen Teber
Source: Georgetown University Medical Center
Contact: Karen Teber – Georgetown University Medical Center
Image: The image is credited to Neuroscience News
Original Research: Closed access. “Evidence for a Spoken Word Lexicon in the Auditory Ventral Stream” by Maximilian Riesenhuber et al., Neurobiology of Language.
Abstract
Evidence for a Spoken Word Lexicon in the Auditory Ventral Stream
Many models of speech processing posit the existence of a neural lexicon that represents whole words. While past work identified a visual lexicon—the Visual Word Form Area—for written words in the ventral visual stream, experimental evidence for an analogous auditory lexicon has been limited. Using functional MRI rapid adaptation techniques, the authors provide evidence that the Auditory Word Form Area in the left anterior superior temporal gyrus houses neural representations that are highly selective for individual spoken words. Familiarization with novel auditory words increases the selectivity of these representations. These results demonstrate parallels in how the brain encodes written and spoken words, revealing convergent processing strategies across the visual and auditory ventral streams.